You are on page 1of 65

HOME

HELP
P

Id ORNL-3391

Contract No. W-7405-eng-26

CHEMISTRY DIVISION

MIXTURES OF METALS WITH MOLTEN SALTS

M. A. Bredig

OAK RIDGE NATIONAL LABORATORY


Oak Ridge, Tennessee
operated by
UNION CARBIDE CORPORATION
f o r the
U.S. ATOMIC ENERGY COMMISSION
W

4’
i

u a-

! c
iii

CONTENTS
i
-
Page
Abstract.......................................................... 1
Introduction...................................................... 1
"Metallic" Metal Solutions........................................ 4
Alkali-Metal-Alkali-Halide Systems.. ......................... 4
Phase Diagrams. ......................................... 4
Physical Properties..................................... 13
Alkaline-Earth-Metal-Halide Systems.. ........................ 23
Phase Diagrams. ......................................... 23
Electrical Conductance.................................. 29
Rare-Earth-Metal-Trihalide Systems. .......................... 30
Phase Diagrams.... ...................................... 32
Electrical Conductivity.. ............................... 37
"Nonmetallic" Metal Solutions..................................... 41
Transition Metals............................................ 43
Posttransition Metals........................................ 44
' ^
S~aly. ..................................................... 51
References................................................... 52
L

,
MIXTURES OF METALS WITH MOLTEN SALTS

M. A. Bredig

A review is presented of various types of solutions of


metals in molten salts, especially in their own molten ha-
lides. With relatively little reference to the older liter-
ature, the progress made in the last 20 years is discussed.
I Roughly, the solutions are classified into two groups: The
i! - -+ metal may retain, to some degree, its metallic properties in
! P
! the solution, or it may lose them through strong interaction
1 % with the salt solvent. The alkali-metal systems are typical
h
! l i examples of the former type, while solutions of cadmium or
1
bismuth represent the second. Equilibrium phase diagram
1 data are presented in detail for many metal-salt systems.
These include critical solution temperatures, that is, tem-
peratures above which metal and salt are miscible in all
proportions. Electrical conductivity is singled out as a
most significant physical property from which conclusions on
the state of the electron in the solution may be drawn. In
the electronically conducting solutions, notably of the al-
kali metals, the electrons may be thought to resemble F
centers in color-center colored crystals. In solutions where
electronic conductance is absent, monomeric, dimeric, and
even more highly polymerized species of the solute metal in a
low valence state must be assumed to occur.

INTRODUCTION

The history of the subject of interactions between metals and their


salts in the molten state, though not well known to most chemists, is at
least 150 years old.' It includes observations made by Davy in 1807 on
the production of deeply colored melts and on difficulties in recovering
metal at the cathode upon electrolysis of molten alkali-metal hydroxides.
Similar phenomena were encountered by chemists who, either in small-scale
laboratory research or in larger-scale industrial production, have been
o r are concerned with the recovery of certain metals by molten-salt elec-
trolysis .* Colored m e t a l "fogs" were recognized by various investigators3
as one important cause of low current yield in the electrolysis of molten
salts. Various theories ranging from colloidal suspensions to "subhalides"

1
F'
2

were proposedto explain t h e observations. One a l s o finds i n the l i t e r - G


ature the statement by N e r n ~ t : ~"For not a single metal do we know a
nonmetallic solvent i n which the metal would dissolve without chemical
interaction and from which t h e pure m e t a l could be recovered simply by
recrystallization ... ." We know today t h a t many such molten-salt sol-
vents do e x i s t . A point of v i e w which appears t o be exactly opposite
t h a t of Nernst, toward which one group of seems t o be inclined,
s t r e s s e s t h e notion t h a t m e t a l atoms may indeed be dissolved as such i n
molten salts, especially when t h e interaction with t h e molten solvent i s
not a strong one. Neither view, i f taken i n t h e extreme, appears t o be
consistent with experimental observation.
Y
The present discussion may profitably be r e s t r i c t e d t o cases i n
which a metal dissolves i n one of i t s own molten halides. S a l t s contain-
ing anions such as n i t r a t e s , sulfates, o r phosphates are l i k e l y t o de-
compose i n reactions with the m e t a l . L i t t l e of a precise nature i s as
y e t known about solutions of one metal i n the molten s a l t of another,
but a few specific cases w i l l be b r i e f l y mentioned. Facts w i l l be
presented which demonstrate t h a t the variety of t r u e solutions of m e t a l s
i n fused salts i s considerable. This includes, a t one extreme, dissolu-
t i o n with r e l a t i v e l y l i t t l e solvent interaction, which might be described
by t h e concept of electrons substituted f o r anions i n cavities, or t h a t
of solvated metallic electrons. A t t h e other extreme, dissolution with
chemical reaction between solvent and solute occurs, as i n t h e formation
of "subhalides," where unusually low ionic valence states of t h e metallic
element are formed. Contrary t o t h e older l i t e r a t u r e 3 as w e l l as t o more
recent statements,8 no evidence whatsoever seems t o e x i s t f o r t h e occur-
rence of t h e colloidal state of metals i n fused salts (with t h e exception
of highly viscous systems such as s i l i c a t e s ) .
It i s convenient t o divide solutions of metals i n t h e i r molten ha-
l i d e s i n t o two main categories:
1. The m e t a l imparts p a r t i a l l y metallic character t o i t s solution
i n t h e salt. The properties of t h e solution r e f l e c t t h e presence of
mobile electrons. These are t h e t r u e metal solutions. They are of par-
t i c u l a r i n t e r e s t because they represent a novel s t a t e of metallic matter.
6
3

W They are somewhat similar t o solutions of metals i n l i q u i d ammonia and


similar solvents, but are distinguished from these largely by the ionic
nature of the solvent (the salt), which, it i s t o be especially noted,
usually contains one of the constituents of t h e metal which it dissolves,
namely the cation. Semiconducting solutions (i.e , those having positive
temperature coefficients of t h e electronic part of t h e conductance) are
included and may, i n fact, represent the majority of examples i n t h i s
class b

2. Strong interaction (i.e., chemical reaction, oxidation-reduction)


occurs between m e t a l and salt. The m e t a l assumes a valence higher than
zero but lower than normal. This c l a s s might be designated "subhalide"
solutions. One might distinguish two subgroups, depending on t h e nature
of the second, t h a t is, the metal-rich phase (solid o r l i q u i d ) that, a t
saturation, i s i n equilibrium with the molten s a l t - r i c h phase:

(a) The reaction goes so far as t o lead t o the c r y s t a l l i z a t i o n of a


s o l i d "subhalide. If
(b) The metallic element i t s e l f (with a s m a l l amount of salt dissolved)
forms t h e second phase. I n this group t h e lower valence state of
t h e metal i s s t a b l e only i n t h e molten solution.

The division between 1 and 2 i s not c l e a r cut, and both mechanisms


of dissolution may describe a single system. The d i s t i n c t i o n i s between
r e l a t i v e l y mobile electrons and electrons which attach themselves to, and
I
become p a r t of, ions t o produce a lower valence state. There are l i k e l y
i
1 - t o be intermediate cases not c l e a r l y defined, and t h e attachment of elec-
!

trons (or "subhalide" formation) may o r may not be a matter of degree


only, rather than a matter of a s t a t i s t i c a l equilibrium between two
d i s t i n c t l y different s t a t e s of t h e electron, attached and unattached.
Also, t h i s may vary with temperature and composition. The variety i n t h e
nature of m e t a l solutions w i l l be i l l u s t r a t e d i n t h e following by a
systematic consideration of a number of examples. A s we are dealing with
mixtures of metals with s a l t s i n t h e l i q u i d s t a t e , we shall a l s o have
occasion t o consider metal-rich phases and the s o l u b i l i t y of salts i n
l i q u i d metals, which i n many cases i s greater than t h a t of the metal i n
t h e molten salt. The discussion will be based largely on measurements of
6) t h e phase behavior and t h e e l e c t r i c a l behavior of these systems.
4

1I "mmC'' METAL SOLUTIONS

The metal+netal halide systems which might be termed "metallic"


metal solutions are mainly those of t h e metals of t h e f i r s t two main
groups of t h e periodic system and of some of t h e rare-earth metals.
E a r l i e r l i t e r a t u r e , that of t h e l a s t century and t h e beginning of t h e
present one, has been reviewed by Cubicciotti,' and a rather complete
review has been published by Ukshe and Bukuneg Many of t h e earlier re-
sults must be considered as of rather limited value, except f o r t h e fact
t h a t they indicated t h e existence of stable mixtures of m e t a l s with
salts, both s o l i d and molten. With a few exceptions, we shall not deal
i n any detail with these older data, many of which were r e l a t i v e l y in-
accurate, or were given erroneous interpretations. We s h a l l confine our-
selves t o t h e more recent work, which i s beginning t o lead t o a far more
s a t i s f a c t o r y explanation of these solutions than was available 20 years
ago.

Alkali-Metal-Alkali- H a l i de Systems

Phase Diagrams

A study of alkaline-earth-metaldkaline-earth-halide systems under-


taken under t h e auspices of the Atomic Energy Research Program i n t h e
United States during World W a r I1 (the "Manhattan FToject")10 and t h e
subsequent investigations of t h e alkali-metal-alkali-halide systems begun
i n t h e early 1950's a t t h e O a k Ridge National Laboratory seem t o have
i n i t i a t e d a new period of considerable i n t e r e s t i n metal-salt solutions.
Except f o r rather demanding experimental conditions, t h e s i t u a t i o n was
believed'' t o be particularly simple with t h e elements of t h e first main
group of t h e periodic system, where complications related t o formation
of a lower valence state of t h e metal were not anticipated. Indeed, t h e
phase diagrams, i n which t h e alkali metal i s one component and one of
i t s halides t h e other, a r e r e l a t i v e l y simple (Figs. 1-6). The principal
data of these diagrams a r e given i n Table 1. The temperature range of

0
L, ,
le c 1 C

UNCLASSIFIED
ORNL-LR-WG. 2 t 8 9 4 A

UNCLASSIFIED t MOLE FRACTION MX t MOLE FRACTION MX


ORNL-LA-DWG. 38872 .9 .8 .7 .6 .5 .4 . 3 .2 .4
'
7

4 200 1200
I I I I I I I I I ' 0
X
COOLING CURVES,THIS WORK
EQUILIBRATION MAB JWJ WTS
HIGH TEMPERATURE RANGE A EOUILIBRATION', H R ~ ~. e '

UNCLASSIFIED
ORNL-LR-DWG 214 DlUM METAL-SODIUM HALIt
SYSTEMS
900
TWO LIQUIDS

900 900

050
800 800

\ SOLID SALT+ LIQUID METAL


700

600

4400 c
I
ONE L I Q U I D

(026'
ONE L i a u i ~
600

4400 Ln

I
10
I
20
I
30
t
40
I I
50 60
mole % K
I
70
I
80
Ml
90 700
-006
t SOLID SALT+ LIQUID METAL

.I .2 .3 .4 .5 .6 .7 .8 .9
N

.I .h .:.k .k ;.
656'
SOLI~SALT+LIOUID

;.
METAL

.b .A 6oo
MOLE %SODIUM M E T A L MOLE FRACTION OF METAL+ MOLE FRACTION OF METAL+
Fig. 1. Potassium Metal-Potassiur n
Halide Systems [J. W. Johnson and M. Fig. 2. Sodium Metal-Sodium Halide Systems, Fig. 3. Sodium Metal-Sodium Halide Systems [M.
A. Bredig, J . P h y s . Chem. 62, 606 (1958) High-Temperature Range [adapted from M. A. A. Bredig and H. R. Bronstein, J . P h y s . Chem. 64,
(reprinted by permission of the copyright Bredig and H. R. Bronstein, J. P h y s . Chem. 65 (1960) (reprintedby permission of the copyright
owner, the American Chemical Society)]. 64, 64 (1960)l. owner, the American Chemical Society)l.
UNCLASSIFIED
ORNL-LR-DW. 34067A

, I I
UNCLASSIFIED
ORNL-lLR-OWG.370

800 I
UNCLASSIFIED
ORNL-LR-DWG. 32909A

MOLE % M E T A L
MOLE X METAL
Fig. 4. Alkali Metal-Alkali Metal
Fluoride Systems [A. S. Dworkin, H. Fig. 5. Cesium Metal-Cesium
R. Bronstein, and M. A. Bredig, 1. Halide Systems [adapted from M. A.
P h y s . Chem. 66, 572 (1962) (reprinted Bredig, H. R. Bronstein, and W. T.
by permission of the copyright owner, Smith, Jr., J. Am. Chem. SOC. 77,
the American Chemical Society)]. 1454 ( 1955)l.
Table 1 Principal Data for Alkali-Metal-Alkali-Halide Phase Diagrams
Monotectic
sat-
Consolute
Metal Salt Tap Phase Comp. (mole $ M) Eutectic Metal
system Compa
MX-M
(m)
(?): (OK) Salt-Rich Metal-
Rich
Temp
(OK)
Comp
(mole $ M) (mole m) (‘E)
Solid Liquid Liquid
LiF 1121 n20 lb 3 1603 40 452
LiCl 883 882b 0.5b 452
LiBr 823 822b 452
LiI 742 741b lb 452
NaF 1268 1263 3b 83 1453 28 10-9 370
NaCl 1073 1068 0.15 2.1 97.7 l353 50 370
NaBr 1020 103.3 0.15 2.9 96.6 1299 52 370
NaI 933 930 1.6 98.6 l306 59 370
KF ll31 1122 4.9 51.7 1177 20 3 x 10-4 337
KC1 1043 1024 0.04C 10.5 75 .o 1063 39 10-9 337
KBr 1007 981 0 .03c 19.o 69.2 1001 44 10-9 337
KI 954 931 13.5 82.5 990 50 10-9 337
RbF 1068 1046 9b 40b 1063 21b 312
RbCl 995 969 18 57 979 37 312
RbBr 965 a a d d globre ub,e 312
RbI 920 888 22 73 907 51 312
CsF 976 d d d d d d 10-3 302
CSCl 918 d d d d d d 10-8 302
CsBr 909 d,a d,a d,a d,a d,a d,a 10-7 302
CSI 899 d d d d d d 10-7 302

%stimate by extrapolation.
bEstimated.
C
By extrapolation from ref 13.
%o miscibility gap.
e
Unstable.
8

special i n t e r e s t i s t h a t near the melting points of t h e salts and above.


The melting points of the a l k a l i metals are very much lower than those
of t h e i r salts and are not f a r above room temperature. The s o l u b i l i t y
of t h e salts i n the l i q u i d metals, which i s considerable i n the higher
temperature range, decreases rapidly with decreasing temperature, so
t h a t the composition of t h e eutectic l i q u i d s i n these systems i s t h a t of
& n o s t pure alkali metal. The mole f r a c t i o n of t h e salt i n these eutec-
t i c s is, with few exceptions, less than loo7, as estimated by extrapol&
t i o n of t h e liquidus curves. Although a eutectic of this nature has
sometimes been known under t h e term "monotectic," this usage w i l l not be
followed here. W e s h a l l use this term1* t o designate the equilibrium
between two l i q u i d s and one s o l i d of a composition not intermediate be-
tween those of t h e two liquids. Thus, i n most alkali-metal-alkali-halide
systems t h e melting point of t h e salt i s lowered by t h e addition of metal
u n t i l t h e monotectic temperature and composition are reached. Above the
"monotectic horizontal" we f i n d t h e region of coexistence of one s o l i d
and one l i q u i d phase, and t h e region of t h e coexistence of two liquids,
one r i c h e r i n salt, the other r i c h e r i n metal. I n t h e alkali-metal sys-
t e m s the compositions of the two l i q u i d phases approach each other
monotonically with increasing temperature u n t i l , a t t h e c r i t i c a l solution
or consolute temperature, they equal each other. A t and above this %em-
perature, onLy one l i q u i d phase e x i s t s under equilibrium conditions.
Below t h e monotectic temperature, s o l i d salt, containing i n s o l i d solu-
t i o n very s m a l l amounts of m e t a l ( w h i c h decrease rapidly with decreasing
temperature),13 i s i n equilibrium with a solution of t h e salt i n l i q u i d
metal .
The general description of t h e phase diagram given here f i t s t h e
lithium,14 sodium,l1*l5 and potassium16 m e t a l systems with each of t h e i r
halides, all of which, with t h e exception of t h e chloride, bromide, and
iodide systems of lithium, are known i n detail. Among these three alkali
m e t a l s t h e potassium systems (Fig. 1) exhibit t h e highest degree of misci-
b i l i t y of the salt with t h e m e t a l i n t h e l i q u i d state. The sodium sys-
tems (Figs. 2 and 3 ) are intermediate, and l i t h i u m metal (Fig. 4 ) shows
t h e least tendency of dl alkali-metal systems t o mix with i t s halides
9

i n t h e l i q u i d phase. The c r i t i c a l solution temperature i n these systems


represents a q u a l i t a t i v e measure of the r e l a t i v e m i s c i b i l i t y and of the
r e l a t i v e deviations from "ideal" solution behavior. T h i s consolute tem-
perature i s 1330°C i n t h e lithium fluoride system, ranges from 1028 t o
1180°C i n the sodium systems, and ranges from 717 t o 904°C i n t h e potas-
sium systems (Table 1). The temperature range i n which two l i q u i d s
coexist i s about 530°C i n t h e l i t h i u m fluoride system, about 200 t o
400°C f o r sodium systems, and only 20 t o 60°C f o r t h e potassium systems.
No precise data appear as yet t o be available f o r ternary a k a l i - m e t a l
systems embodying two o r more cations with one anion (besides t h e elec-
tron), o r several anions with one cation.
It i s evident that t h e c e s i w e s i u m - h a l i d e systems17 a r e qualita-
t i v e l y d i f f e r e n t from the majority of the alkali-metal-alkali-halide sys-
t e m s , i n t h a t t h e l i q u i d s a r e miscible i n a l l proportions (i.e., the
consolute temperature l i e s below t h e liquidus l i n e ) (Fig. 5 ) . The l i q u i -
dus curve, depicting t h e temperatures f o r t h e s o l i d i f i c a t i o n of t h e ce-
sium halide from t h e l i q u i d mixture, descends without discontinuity
(except t h a t t h e r e ought t o be a slight kink a t t h e transformation tem-
perature of CsC1) from the melting point of the pure salt t o t h e eutectic
point, which i s almost i d e n t i c a l i n melting temperature (30°C)and compo-
s i t i o n with t h e pure metal.* The four rubidium phase diagrmslg show, as
might be expected, a behavior intermediate between the potassium and t h e
cesium systems (Fig. 6). The temperature range of only p a r t i a l miscibil-
i t y i s l e s s than 20°C i n the fluoride, chloride, and iodide systems, and
j : t h e m i s c i b i l i t y gap i s absent i n t h e bromide system. It i s i n t e r e s t i n g
t o note (Fig. 7) t h a t the trend, with increasing atomic nuniber, size, or
p o l a r i z a b i l i t y of t h e halide ion, toward greater m i s c i b i l i t y as expressed
by the value of the c r i t i c a l solution temperature [Tc = U80, 1080, and
1026°C f o r Na(F, C1, B r ) ; 904, 790, and 728°C f o r K(F, Cl, B r ) ; and 790,
706, and < 655°C f o r Rb(F, C1, B r ) respectively] is, i n t h e iodide systems
*Delimarskii and Markov ( r e f 18), i n Table 29, page 211, "Solubility
of Metals i n Fused Salts," misstate t h e case of t h e cesium systems, by
erroneously taking t h e metal concentration of t h e solution i n equilibrium
with s o l i d salt f o r a metal concentration of a s a l t - r i c h phase i n equilib-
rium with l i q u i d cesium metal. Many other metal s o l u b i l i t i e s i n t h i s
t a b l e a r e much outdated and quite i n error.
10 c l

UNCLASSIFIED
ORNL-LR-WIG. 6707

'200 t 1 ' 1 ' 1 ' 1 ' -

Fig. 7. Critical Solution Temperature in Alkali


Metal-Alkali Metal Halide Systems vs Molar Refrac- L,OC

tion of the Gaseous Halide Ion. 900

800
1 F

700 I
600
0
i
I I
4
I t
R
8
I
Rb

' I
42
- (cm3 mole-')
' I
46
'

UNCLASSIFIED
ORNL- LR-DWG. 701
5000 I I I I I I I I I

4000
d
- 3000
-
c
0

%
s - I
-
c
L

t
k 2000

4000

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.0 0.9 0


4;
Fig. 8. Excess Free Energy of Mixing as Function
of Volume Fraction of Metal.

- I
11

of these three alkali metals, e i t h e r reversed [Na(I):1033'] o r consider-


ably diminished [K(I):708' and Rb(I):634'], with data lacking f o r the
l i t h i u m systems and f o r Cs-CsBr. Although it seem reasonable that the
solution behavior and Tc might be r e l a t e d t o the s i z e and p o l a r i z a b i l i t y
of t h e anions, no simple quantitative relationship t o these anion proper-
t i e s has been deduced.
Besides t h e value of the minimum temperature of complete m i s c i b i l i t y
(Tc), t h e ( " c r i t i c a l " ) composition a t the consolute temperature i s of
interest. The metal content of the " c r i t i c a l " solution i s seen, i n
Table 1, t o increase f r o m t h e fluoride t o the iodide f o r t h e sodium, po-
tassium, and rubidium systems; t h e same i s t r u e f o r the compositions of
t h e i n f l e c t i o n on t h e solidus curve of t h e cesium systems, which i s some-
.
w h a t representative of t h e "submerged" (i e., unstable) c r i t i c a l solu-
tions (28, 47, and 65 mole $ cesium metal f o r CsF, CsC1, and C s I solution
with cesium respectively). It has been pointed out t h a t the correspond-
ing " c r i t i c a l " volume fractions approach t h e value 0.50 f o r a l l systems,
indicating t h e quite plausible significance of t h e molar volume i n the
solution behavior .I6
The a c t i v i t y coefficients of t h e salts, calculated from t h e liquidus
curves f o r t h e precipitation of s o l i d salts from t h e m e l t s , increase from
unity with increasing metal concentration, corresponding t o positive devi-
ations of t h e salt component from Raoult's law. This i s usual i n systems
exhibiting (or having a tendency t o exhibit ) liquid-liquid m i s c i b i l i t y
gaps, when no strong ("chemical") interactions such as i n t h e cadmium-
cadmium halide systems (cf. below) take place. I n Fig. 8 are plotted
values of RT ln ySat
vs % e t d f o r the systems K-KBr, Rb-RbBr, and
Cs-CsI, where i s t h e volume f r a c t i o n of the metal.
metal
The temperature dependence of t h e s o l u b i l i t y of the s o l i d salts i n
t h e l i q u i d m e t a l s i s another feature of i n t e r e s t .
An unusually low value
of t h e p a r t i a l molar heat of solution i s derived from t h e temperature
dependence of t h e s o l u b i l i t y of t h e s o l i d salts i n the l i q u i d m e t a l s f o r
the fluorides of cesium (10 kcal/mole) and potassium (l3 kcal/mole) i n
comparison with the other halides of these metals.
No explanation of these phenomena which i s f r e e of a r b i t r a r y assump-
t i o n s has been proposed, although r a t h e r i n t e r e s t i n g speculations1'* 2o
lead t o a crude but i n s t r u c t i v e picture of the solution behavior of these
systems.
Pitzer21 suggests t h a t the excess f r e e energy of mixing i s associated
with the conversion of t h e metallic s t a t e of binding of the metal elec-
trons t o an ionic type of binding and t h a t t h e mixing of electrons with
halide ions occurs with l i t t l e excess f r e e energy, i f any. He obtains
fair agreement between experimental excess f r e e energies derived from
the phase diagram and a t h e o r e t i c a l caJmLlation of t h e difference i n
energy of a metallic and ionic model of an alkali metal.
A n i n s t r u c t i v e view may be obtained according t o Blander22 by dividing
the process f o r the dissolution of metal i n t o the two steps:

M(1iquid) +* M(gas) ,
M(gas) -+ in d i l u t e solution) .
The first s t e p i s simply t h e vaporization of the metal f o r which data are
available.23 The thermodynamics r e l a t e d t o t h e second step, which i s t h e
dissolution of t h e gas molecules, i s discussed elsewhere.24 The i n f l u -
ence of the f i r s t s t e p may be i l l u s t r a t e d by the r e l a t i v e values of RT
In p given i n Table 2, where p i s the p a r t i a l pressure of the metal mono-
mer i n t h e vapor i n atmospheres. Aside from differences i n t h e second
process, the influence of the vaporization s t e p would tend toward an in-
creasing r e l a t i v e s o l u b i l i t y of metal i n salts i n the order of increasing
vapor pressure, L i < Na < K < Rb < C s .
Where Henry's law applies, t h e second process i s b e s t described i n
terms of the Henry's law constant, %,

where csol i s the concentration of metal i n solution i n moles/cm3 and


c i s t h e concentration of metal i n t h e gas phase i n the same units.
gas
The f r e e energy of solution (ac = -RT I n Kc) i s the sum of t h e f r e e
e n e r w of forming a hole the s i z e of t h e metal atom (about 10 2 5 kcal/
mole) and the f r e e energy of interaction of the metal atom with t h e salt.
W Table 2. Miscibility and Metal Vaporization Data f o r t h e
Alkali- Met al-Alkali- H a l i de Systems

RT ~n
I
p monomer)
(kcal mole)
( a t looo'~)
Metal bp
(OK)
Range of Values of Tc
(OK)

Li -13.8 1604 1603

Na -3.6 ll63 1303-1453

K 4.8 1039 981-ll77

Rb 43.3 974 907-1063

cs 43.6 958 (<850 t o 4 5 0 "submerged")

A s an example, one might consider t h e dissolution of sodium i n a hypo-


t h e t i c a l sodium salt having a molar volume of about 50 cm3 a t the boiling
point (1163'K) t o form a 1mole $ solution. These parameters are not
very different from those f o r r e a l salts. The free-energy change, &,
f o r this process, i f Henry's l a w holds even approximately, i s -7 kcal/mole,
and t h e interaction energy of a metal atom with t h e solution would have
t o be about -17 5 5 kcal/mole. Since t h e s o l u b i l i t i e s of sodium are
higher than 1mole $, the postulate of t h e dissolution of metal atoms with
weak i n t e r a c t i o n i s seen t o be improbable. That t h e dissolution of a l k a l i
m e t a l s i n molten salts i s influenced greatly by t h e same factors influ-
encing the vaporization i s shown by t h e comparison i n Table 2 of the boil-
ing points of t h e m e t a l s and t h e consolute temperatures. This is, of
course, only suggestive, and much more data and a more complete quantita-
t i v e analysis are necessary f o r any d e f i n i t e conclusions.

Physical Properties

Optical Observations. - Aside from the early qualitative findings


about t h e deep coloration of molten a l k a l i halides containing excess
metal, there appears t o be, thus far, only one b r i e f study of a semiquan-
t i t a t i v e nature. M 0 l l w 0 ~measured
~ t h e v i s i b l e and near-infrared absorp-
t i o n spectra of m e l t s of alkali-halide c r y s t a l s colored by exposure t o
14

alkali-metal vapor. He found broad absorption m a x i m a a t 790 mp f o r all


three sodium salts studied (NaCl, NaBr, and N a I ) and a t 980 m + ~f o r the
corresponding potassium salts. I n contrast t o this, t h e position of t h e
-
absorption maximum i n t h e s o l i d salts varies with t h e anion, or t h e l a t -
t i c e parameter. Mollwo believed t h a t t h e absorption bands of the m e l t s
could be considered as resonance l i n e s of t h e metal atoms, displaced and
broadened by intermolecular fields. However, it seems e n t i r e l y possible
t o ascribe t h e absorption spectrum t o an electron whose wave function i s
not confined t o one metal core (M) but spread out over several M+ ions
as i n an F center, but without t h e decisive influence, upon i t s
energy, of a r i g i d crystal. l a t t i c e .
ELectrical Properties. -
I n a very f e w r a t h e r crude experiments,
M 0 l l w 0 ~found
~ no directional transport of the colored cloud from a
melting, F-center colored c r y s t a l . He considered this as a confirmtion
of the o p t i c a l finding, which indicated t h e presence of uncharged (i.e.,
nondissociated) m e t a l atoms.26 This conclusion does not s e e m t o be con-
firmed by l a t e r observations.
Bronstein and Bredig2 7 s measured t h e conductance of a l k a l i - m e t a l -
alkali-metal-salt mixtures and demonstrated very considerable transport
of e l e c t r i c i t y by electrons i n solution. The stainless s t e e l apparatus
f o r the determination of t h e specific e l e c t r i c a l conductivity, which
they describe i n detail, included a synthetic-sapphire c a p i l l a r y dip c e l l
which was found t o be e n t i r e l y i n e r t t o most alkali-metal-halide mixtures.
ALL of t h e seven sodium and potassium systems investigated (the tempera-
t u r e s of t h e Na-NaF system were not accessible) had i n common a r i s e i n
t h e specific conductivity of the salt melt on addition of m e t a l . 2 7 * 28

T h i s r i s e , however, varied greatly i n extent and nature from system t o


system. Figures 9, 10, and 11 a r e ty-pical examples. I n t h e sodium sys-
tems t h e rapid i n i t i a l r i s e i n conductivity i s c h a r a c t e r i s t i c a l l y slowed,
and only i n cases of s u f f i c i e n t s o l u b i l i t y (Na-NaBr and Na-NaI a t higher
temperatures) i s it f i n a l l y accelerated as t h e s o l u b i l i t y l i m i t i n con-
centration i s approached. I n t h e potassium solutions, on t h e other hand,
t h e r i s e i s monotonically accelerated with increasing m e t a l concentra-
tion. I n solutions of potassium i n K I , a specific conductivity of 600
ohm-’ cm-I, that is, 400 times t h a t of t h e pure molten salt, was measured
15

a t 42 mole 76 potassium, indicating an electronic contribution t o the con-


ductivity of t h i s solution of more than 99.576. The potassium fluoride
solutions a r e distinguished from those i n t h e iodide by a f a r slower
i n i t i a l r i s e i n conductivity, with the chloride and bromide melts being
intermediate. Bronstein and Bredig define a c h a r a c t e r i s t i c or apparent
equivalent conductance of t h e dissolved m e t a l by the equation

UNCLASSIFIE0
ORNL-LR-DWG. 23048A

6P '845'

-
2l
NO NoCl

Na-NaBr
2-

0 2I I 4l ~6 l 8~ 0l ,2 l 4, l 6, 8l , 4l ,
0 l , ~ o
SODIUM METAL (mole %) POTASSIUM METAL (mole %)

Fig. 9. Specific Conductivity of Alkali Metal


Solutions in Alkali Halides [H. R. Bronstein and M.
A. Bredig, 1. Am. Chem. SOC. 80, 2080 (1958) (re-
printed by permission of the copyright owner, the
American Chemical Society)].
16

UNCLASSIFIED
ORNL-LR- DWG.54649
I I I I I I

J
Fig. 10. Specific Conductivity of Solutions of
Sodium Metal in Sodium iodide [H. R. Bronsteiii and
M. A. Bredig, J . P h y s . Chem. 65, 1221 (1961) (re-
printed by permission of the copyright owner, the
American Chemical Society)].

0
I I
2
I
4 6
I I I
S i 0 4 2
I
4
MOLE PER CENT SODIUM METAL

UNCLASSIFIED
ORNL-LR-DWG. 54650

48 - -

46 - -

-
s
0
- 42- -

-
a
2
-

I I I I I
0 2 4 6 8 4 0 4 2
Ir' where Asoh are t h e equivalent conductances of t h e solution
and
and salt, and NM i s t h e equivalent f r a c t i o n of metal. The term % repre-
sents t h e change of t h e conductance caused by t h e addition of one equiva-
l e n t of m e t a l t o an mount of salt contained i n t h e solution of t h e given
composition which i s between two electrode plates 1 cm apart.
i Figures 1 2 and 13 show, i n a more s t r i k i n g form than Figs. 9, 10, and
11, t h e t y p i c a l difference i n the behavior of sodium and potassium. While
i n t h e potassium systems t h e apparent equivalent conductance of t h e metal
solute, %, -
rises monotonically and with increased r a t e with increasing
m e t a l content, it -
drops rapidly i n t h e sodium solutions toward a minirmun,
which obviously would be followed by a rise i f greater s o l u b i l i t y would
permit higher concentrations of metal t o be dissolved.
These c h a r a c t e r i s t i c changes of % w i t h m e t a l concentration m y be
interpreted as follows: A t i n f i n i t e dilution of the m e t a l , i n both the
sodium and potassium solutions, electrons are i n a s t a t e i n which they
can contribute t o the carrying of current. Rice,29 assuming a "random
walk" of t h e electrons, has performed t h e o r e t i c a l calculations f o r this
s t a t e a t infinite dilution which are correct within an order of magnitude
of t h e observed conductivity. For t h e potassium solutions, Rice's theory
predicts an increase of & w i t h m e t a l concentration. Bronstein and Bredig
suggest t h a t electron o r b i t a l overlap plays a r o l e i n this increase, and
doubtlessly this becomes important a t higher m e t a l concentrations. How-
ever, i n t h e sodium solutions, another important factor appears t o be
present t h a t actually removes electrons from t h e conduction process with
increasing m e t a l concentration. The trapping of electrons i n pairs t o
form diatomic molecules of Na2 was suggested27 t o r a t i o n a l i z e this dif-
ference. !The f a c t t h a t Na2 i s r e l a t i v e l y more stable i n t h e vapor than
i
K2 (the heats of dissociation are 17.5 and 11.8 kcal/mole respectively)
suggests t h a t this i s reasonable. Measurements a t higher temperature i n
t h e sodium-Fontaining systems, i f t h i s hypothesis were true, would lead
t o behavior more l i k e the potassium-containing systems because of t h e
greater dissociation of Na2. In t h e l i t h i u m systems a s t i l l greater de-
gree of pairing should occur. This, r a t h e r than a short relaxation time,30
iw
i
might be t h e reason f o r t h e absence of electron spin resonance i n solutions
of lithium i n molten lithium iodide.
18

UNCLASSIFIED
ORNL-LR-OWG 23047A

-
c)
1
40-
No in NH3
0
T
- at -335'

a -
CT

Fig. 12. Equivalent Conductance of


Alkali Metal Dissolved in Alkali Halides
[H. R. Bronstein and M. A. Bredig, J .
Am. Chem. SOC. 80, 2080 (1958) (re-
printed by permission of the copyright
owner, the American Chemical Society)].

0 2 4 6 0 40 42 44
METAL (mole %)

Fig. 13. Equivalent Conductance of


Alkali Metal Dissolved in Alkali Halides
[H. R. Bronstein and M. A. Bredig, J .
Phys. Chem. 65, 1223 (1961) (reprinted
by permission of the copyright owner,
the American Chemical Society)].
Another i n t e r e s t i n g aspect i s the v a r i r t i o n of 4(i.e., n, extrapo-
l a t e d t o infinite dilution) with variation of e i t h e r the anion or t h e
metal ion. Table 3 gives a summary of the equivalent molar conductances,
4,of sodium and potassium metals i n i n f i n i t e l y d i l u t e solution i n most
of t h e i r molten halides. These values were obtained by a short extrapo-
lation, t o zero m e t a l concentration, of the curves for vs m e t a l con-
centration.
Two d e f i n i t e trends may be seen i n these data: F i r s t , the contribu-
t i o n t o the conductivity by t h e metal solute increases greatly i n going
from t h e fluoride t o the iodide systems, t h a t is, with atomic number, or
size, of t h e halide ion. Second, it decreases i n going from sodium t o
potassium (i.e., apparently, with increasing atomic number, or size, of
the metal ion). The former trend has been attributed27* t o the increase
i i n t h e p o l a r i z a b i l i t y of t h e halide ion with increasing size, which i s
thought t o f a c i l i t a t e transmission of an electron from one cation, or
group of cations, t o another. This i s reminiscent of t h e findings by
Taube and others t h a t t h e r a t e constant f o r exchange of an electron be-
tween a complexed cation such as [Cr(NH3)5X12+ or [Cr(H20)5XI2+ and hy-
drated Cr2+ ions i s very highly dependent on the nature of t h e ligand
halide ion, Y, presumably on i t s polarizability. However, while t h e ef-
f e c t i n aqueous solution shows a very high power dependence on the polari-
z a b i l i t y of T (the bimolecular rate constant f o r electron exchange be-
tween [Cr(NH3)5Xl2+ and C r 2 + changes from 2.7 x lr4t o 5.1 x 1r2t o 0.32
t o 5.5 f o r X = F, C1, Br, and I respectively), i n t h e metal-salt m e l t s
I * the apparent dependence of L$ on p o l a r i z a b i l i t y of X i s of a lower power.
i
Quite significantly, the iodide systems exhibit a weakening of t h e trend
established by t h e fluoride, chloride, and bromide systems. This i s remi-
niscent of a similar change i n t h e trend for t h e c r i t i c a l solution temper-
atures mentioned above, a s i m i l a r i t y which may not be merely coincidental.
O f considerable i n t e r e s t would be a discussion of t h e temperature
dependence of t h e electronic part of t h e conduction i n i n f i n i t e l y d i l u t e
solution. Unfortunately, the data are, by t h e i r very nature (i.e., small
differences of two large numbers divided by a small concentration value),
not nearly accurate enough for t h i s purpose, except t o show t h a t the tem-
_i_
W perature dependence a t i n f i n i t e dilution i s very s m a l l . A t higher metal
L
!

20

Table 3. Molar Conductance 4


of SodLum and Potassium i n I n f i n i t e l y
Dilute Solution i n Their Molten Halides a t 9OOOC (0htd-l cm2 moleg1)

Crystal Radius e1 Br I
of cation (A)

Na 0.95 6000

K 1.33 800 2800 6,000 8,100

. . . . . . . . . . . . . . . . . e
(820O C )
....
(870OC)
b e

Molar r e f r a c t i o n 2.5 9 1 207 19


of gaseous
anion
(cm3 moleo1)
Cube of c r y s t a l 2.52 5 094 7.43 9 095
radius (A3)

%stimated.

4 5 6
REDUCED TEMPERATURE, r/e
Fig. 14. Electron Mobility in Alkali Halide Crystals
[A. Smakula, Nu&. Ges. Wiss. GBttingen, 11 1(4), 55
(1934)l.
21

concentrations t h e temperature coefficient (positive) of the m e t a l solute


conductivity appears t o depend both on metal concentration and temperature
(cf. Fig. 13). T h i s might be attributed27*2 8 t o the equilibrium between
associated species such as Na2 and K2 and electrons, 2 e + 2e' F= K2. At
s t i l l higher metal contents, where the l i q u i d i s e s s e n t i a l l y a solution
of salt i n metal, t h e temperature coefficient appears t o become negative,
as expected f o r metals.
The very s m a l l temperature dependence of the metal solute conductivity
a t i n f i n i t e dilution, 4, s i g n i f i c a n t l y distinguishes t h e melts from t h e
F-center colored c r y s t a l s studied by Smakula.31 I n t h e c r y s t a l s t h e rela-
t i v e l y large activation energy of conductance by F-center electrons i s
d i r e c t l y r e l a t e d t o the thermal. motions of the ions (i.e., t h e character-
i s t i c temperature of the crystal. as derived from the specific heat). Here,
the l a r g e differences i n t h e electron mobility f o r different halide crys-
t a l s of the same a l k a l i metal are wiped out by the use of a reduced tem-
perature scale, Tabs/@ (Fig. 14). A similar relationship using, f o r ex-
ample, t h e melting point of the s a l t obviously (cf. Fig. 13) does not
exist for &, which seems nearly constant with temperature and corresponds
t o an electron mobility several orders of magnitude greater than that i n
the c r y s t a l . It appears, then,that when t h e thermal motion of the ions
and t h e electron mobility a r e as high as they are i n the m e l t , t h e re-
maining individual differences i n 4 may be expected t o be due t o other
aspects of the conduction mechanism, such as t h e p o l a r i z a b i l i t y of t h e
anions which a c t as negative p o t e n t i a l barriers, as discussed above.
I n both t h e c r y s t a l s and the melts, t h e electrons i n the sodium sys-
,
t e m s exhibit higher mobility than i n the potassium systems. T h i s does
not seem t o be w e l l understood a t present, although a connection with t h e
smaller i n t e r i o n i c distances i n t h e sodium case may e x i s t .
We have seen t h a t it i s possible t o prepare binary mixtures of many
alkali metals with t h e i r molten halides i n all proportions of t h e two
components, salt and m e t a l . Compared with t h e study of t h e conductivity
of t h e s a l t - r i c h solutions, only a start has been made i n measuring t h e
conductivity of metal-rich solutions. Figure 15 shows the whole concen-
t r a t i o n range f o r several potassimpotassium-halide systems and demon-
,1 - w s t r a t e s the l a r g e deviation from additive behavior. Bronstein, Dworkin,
' -
22
!
,7, I I , I I
UNCLASSIFIED
pRNL;LR-DyG.5726:

Fig. 15. Specific Conductivity of Solutions of


Potassium Halides in Potassium Metal [H. R.
Bronstein, A. S. Dworkin, and M. A. Bredig, J .
Chern. Phys. 37, 677 (1962) (reprinted by permission
of the copyright owner, the American Institute of
Physics)].

MOLE PER CENT POTASSIUM

and Bredig32 found that up t o 20 mole KI the r e s i s t i v i t y of potassium


m e t a l increased l i n e a r l y with increasing salt concentration.
Subsequent work33 showed that t h e proportionality factor a i n t h e -
equation f o r t h e specific r e s i s t i v i t y , p (pohm-cm) = aNKI + b, increases
with t h e s i z e of t h e anion, from 420 f o r KF (700'), t o 740 f o r XBr (74'),
t o 920 pohm-cm f o r K I (700'). A n appropriate explanation appears t o be
t h e increase i n the cross section of the l a r g e r anions f o r s c a t t e r i n g
moving electrons. An a l t e r n a t i v e explanation, that of t h e exclusion of
conducting volume by t h e introduction of insulator (salt) molecules, i s
e n t i r e l y unsatisfactory: It i s t r u e t h a t the equivalent conductivities,
t h a t is, t h e conductivities of 1mole of solution of salt i n m e t a l , of
KI, KBr, or KF i n potassium, f a l l e s s e n t i a l l y on t h e same curve when
plotted against volume fraction. However, t h e conductivity drops far
more rapidly, i n i t i a l l y by a f a c t o r of 5 t o 10, than t h e exclusion of
conducting volume would require, namely, a t 700°, from approximately
990,000 0hm-l cm2 moleo1 f o r pure potassium m e t a l t o 620,000 a t a volume
23

f r a c t i o n of potassium of 0.95, t h a t is, (990,000 - 620,000)/0.05 x


990,000 = 370,000/49,500 or 7.5 times f a s t e r . It i s not unexpected t h a t
the macroscopic picture of exclusion of conducting volume does not work
i n this s i t u a t i o n where the mean f r e e path of the electrons i s large com-
pared with t h e excluding p a r t i c l e size, which makes the s c a t t e r i n g process
t h e all-important feature .
Alkaline-Earth-Metal-Halide Systems

Phase Diagrams

The older l i t e r a t u r e 1 t 3 contains a r a t h e r bewildering variety of


claims f o r t h e existence of extensive s o l u b i l i t y of alkaline-earth metals
i n t h e i r molten halides and the existence of subhalides of t h e formula
MX o r M2X2 (similar t o Hg2X2). The assumption of s o l i d alkaline-earth
subhalides, f o r example, the claim by Wshler and R ~ d e w a l dregarding
~~
the existence of a s o l i d CaC1, based on chemical analysis of deep-red
c r y s t a l s prepared by fising together calcium and C a C l 2 , has not been borne
out by l a t e r investigations of Eastman, Cubicciotti, and Thurmond10*35 and
of Bredig and Johnson.36 Considerable s o l u b i l i t y of t h e metals i n t h e i r
salts and some s o l u b i l i t y of the salts i n the l i q u i d metal were found, but
i
i
l no indication of s o l i d reaction products, subhalides (MX). A more recent
suggestion37*38 of a layer-type crystal. structure f o r a preparation pre-
s u e d t o be "CaC1" turned out soon thereafter39 t o be erroneous; the crys-
t a l s investigated were those of the ternary compound CaC12TaH2, or CaClH.
Only a start has actually been m d e t o obtain truly accurate phase
diagrams of t h e a3lraline-earth-metaL-halide systems . Schgfer and NiklasIO
i n 1952 very b r i e f l y reported a study of t h e system Ba-BaCl2, which i s
characterized by a s o l u b i l i t y of 15 mole $ barium i n molten BaCl2 and of
5 mole $ BaC12 i n l i q u i d barium metal a t t h e monotectic temperature of
878"C, and by a r e l a t i v e l y low consolute temperature of 10IO°C, only 50'
above t h e melting point of the salt. These r e s u l t s , confirmed by Peterson
# ~ ~ 16), differed greatly from those obtained ear-
and H i ~ & e b e i n ~ l(Fig.
l i e r by Eastman, Cubicciotti, and T h ~ o n d l which gave l i t t l e , i f any,
~ t ~ ~
I
1 24

1
I

I
I
I

I:
700 - " 0

a+#
0 0 0 n n

(I + L,'
Y,
-
I I I I I I I I I I I
0 40 20 30 40 50 60 70 80 90 f00
BaClp MOLE PER CENT Ea Ea

Fig. 16. Barium Metal-Barium Chloride System (J. A.


Hinkebein, Ph.D. Thesis, Iowa State College, 1958).

indication of complete m i s c i b i l i t y a t higher temperatures. The difference


i s due t o faults of t h e e a r l i e r technique i n determining t h e temperature-
concentration area of the coexistence of two l i q u i d s by analysis of
quenched m e l t s . As emphasized by Bredig and c o - ~ o r k e r s ~ and
~ * by
~ ~ * ~ ~
others,40 it i s i n general impossible t o preserve, by quenching, t h e
equilibrium concentration of a l i q u i d phase saturated with respect t o
another l i q u i d phase, because of rapid precipitation and segregation of
t h e second l i q u i d phase. What one may obtain, instead, i s a phase re-
sembling t h e monotectic l i q u i d i n composition. Another occasional source
of e r r o r i s t h e contraction of t h e salt phase on freezing, which produces
a sump hole and/or cracks which f i l l with excess l i q u i d m e t a l . To avoid
these errors it i s necessary t o r e s o r t t o e n t i r e l y different techniques
such as thermal o r d i f f e r e n t i a l thermal analysis, t h a t is, t h e taking of
cooling curves, as SchEfer and Niklas did, o r t o a process of separating
a sample of each l i q u i d phase a t t h e equilibration temperature as de-
scribed by Bronstein and Bredig.27 A s t i l l different method, which relies
on diff'usion of t h e metal i n t o t h e salt contained i n a separate compart-
ment of a capsule,44 i s subject t o some doubt, as t h e wetting character-
i s t i c s of the salt and m e t a l and t h e creeping of l i q u i d m e t a l along t h e
i i
25
I

w inside walls of t h e m e t a l containers may a l s o lead t o d i s t o r t e d results,


aside from t h e inconvenience of long periods of time f o r equilibration.
Besides t h e measurements on t h e Ba-BaC12 system, the results a r e
believed t o be reasonably accurate f o r t h e Ca-CaF2 system reported by
Rogers, Tomlinson, and R i c h a r d ~ o n ; ~ f' o r t h e Ca-CaC12 system reported by
Hinkebein and P e t e r s ~ n ~ and
l r ~by
~ other^;^^^^^ and f o r the higher tem-
perature range of the Ca-CaBr2 and C a - C a I 2 reported by Staffansson,
Tomlinson, and Richardson. 4 4
A low s o l u b i l i t y of magnesium i n molten MgC12 was observed, 0.90
and 0.30 mole '$ (Rogers, Tomlinson, and R i c h a r d ~ o n ~ both
~),
a t 80O'C. A comparison (cf. Tables 1 and 4 ) of the s o l u b i l i t y of t h e
various alkaline-earth metals i n t h e i r molten chlorides near t h e melting
points of the salts reveals a trend somewhat similar t o that found i n
t h e a l k a l i - m e t a l systems, t h a t is, a rapid increase i n s o l u b i l i t y with
atomic number of t h e metal: Mg-MgCl2, < 0.5; Ca-CaCl2, 2.7; Sr-SrCl2,
5.5; Ba-BaC12, 20 mole $.
I n t h e Ca-CaF2 system a r a t h e r high miscibility, t h a t is, a rather
small temperature range (30') f o r coexistence of two liquids, was found,45
with a c r i t i c a l solution temperature of l322'C, -
almost 100' below t h e
melting point of t h e salt (Fig. 17). The c r i t i c a l solution temperatures
i n t h e other three calcium-halide systems a r e very similar, y e t t h e tem-
perature range of only p a r t i a l m i s c i b i l i t y i n these systems covers 500'.
This, of course, i s caused by t h e r e l a t i v e l y much lower melting points
of C a C l 2 , C a r 2 , and Ca12 (772, 742, and 78OoC respectively), as compared
F with 1418' f o r CaF2 with an e n t i r e l y d i f f e r e n t c r y s t a l structure. From
a comparison of t h e melting-point depression of C a F 2 produced by calcium
with that produced by NaF, Rogers, Tomlinson, and Richardson45 concluded
t h a t t h e metal dissolved e i t h e r as calcium atoms or as Caz2+ molecule
ions, namely, according t o e i t h e r C a + Cao o r Ca + Ca2+ 4 Caz2+, both
giving a cryoscopic number, n, of unity, or approximately unity, i n t h e
-
Raoul-GVan't Hoff formula &I2 = (Tm T) = (RT2/%)*m2, where N2 i s the
mole f r a c t i o n of calcium metal. I n both cases, s o l i d solution was as-
sumed not t o occur, and it was concluded t h a t t h e reaction Ca + Ca2+ +
2Ca+ did not take place. However, i n this case, the %in the formula
salt- Eutectic Consolute
Metal Salt Rase Comp (mole $ M) Metal
System mp Temp Comp Temp Comp
MX2-M Temp Salt-Rich
(mole 46 M)
(OK)
Metal- (OK) (mole M)
(m21 (OK) Rich
(OK)

Solid Liquid Liquid

MgCl2 987 -987 ? 0.2 -923 -100 923

C@2 1691 1563 a 25.5 67 1094 98.6 1595 k 5 45 U O

CaC12 1045 1093 2.70 99.5 1033 2 .o 1610 k 5 62 lll0

Ca33r2 1015 1100 2 .3b 99.6 1000 3.0 ? 1610 k 5 64 U O

CaI2 1053 1104 3 .8b 99.7 1033 2.0 ? 1650 k 5 74 lll0

SrC12 U.45 lll2 a 5 05 ? ? ? ? ? 1044

BaC12 1235 1163 C 15.O 95 985 -99 1290 50 1002

a
Considerable metal s o l u b i l i t y i n ( f l u o r i t e type o f ) s o l i d i s l i k e l y .
bSee r e f 46.
C
Considerable metal s o l u b i l i t y i n high-temperature c r y s t a l form of BaCl2 ( f l u o r i t e type ?) is
likely .
iu UNCLASSIFIED
ORNL-LR-DWG. 70746

I
I

Fig. 17. Calcium Metal-Calcium Fluoride System 5


(P. S. Rogers, J. W. Tomlinson, and F. D. Richard- ((00
son, p 919 in Physical Chemistry of Process Metal- k!
0 DIFFERENTIAL
lurgy, Interscience, New York, 1961).
0 SOLUBILITY

0 50 4 00
COMPOSITION ( mole % Ca 1

would become approximately 10 kcal, quite different from the calorimetric


value of 7.1 ( N a y l ~ r ~ which
~ ) , i s hardly i n doubt. It seems, then, more
reasonable t o assume s o l i d s o l u b i l i t y a c t u a l l y t o occur both with NaF and
Ca (as CaF??). MollW0,49 i n measurements on F centers, found the solu-
b i l i t y of calcium i n s o l i d CaF2 a t 1300°C t o be approximately 30 mole $,
which might w e l l be too l a r g e by as much as a f a c t o r of 10. With t h e
value of 7.1 kcal f o r AHm, t h e modified formula

- N2 (Solid)]
would leave t h e p o s s i b i l i t y open f o r a cryoscopic n l a r g e r than 1,
possibly 2, a t low m e t a l concentration. T h i s cryoscopic n = 2 would be
i n agreement with t h e findings i n t h e corresponding chloride system,
CaCl2-Ca, where Dworkin, Bronstein, and Bredigso found n a l s o t o be only
s l i g h t l y less than 2. Here, however, no s o l i d solution was assumed, the
c r y s t a l s t r u c t u r e being quite different, with s i x f o l d r a t h e r than eight-
f o l d coordination of anions around cations. I n both cases the reaction
would then seem t o be Ca + Ca2+ + 2e-, as also suggested by t h e elec-

I
U t r i c a l conductivity (discussed below), with a possible, though much l e s s
?i
28

plausible, additional. dissolution mechanism, Ca + Ca2+ -+ 2Ca+, not ex-


cluded at this stage, and with increasing formation of (Ca2)2+ at higher
metal concentration.
In the Sr-SrC12 system, Staffans~on~~ found n closely equal to 1,
assuming a heat of fusion of 4.1 2 0.6 kcal. Dworkin, Bronstein, and
Bredigso questioned this latter heat value as possibly being too low.
However, a recent calorimetric determination (Dworkin and Bredig'l) essen-
tially confirmed the low value, which actually gives fair agreement in the
rather low entropies of fusion of CaF2 and SrCl2 (4.2 and 3.4 e.u.) of
similar, fluorite-type crystal structure. Mollwo found the solubility of
strontium in solid SrC12 to be high, similar to that of calcium in CaF2.49
As in the case of CaF2-Ca, the existence of solid solution in the fluorite
type of structure could lead, then, to an interpretation of the melting-
point depression in terms of n w 2, corresponding to Sr &-+ -, 2e-
--+

(and/or, less likely, Sr + Sr2+ + 2Sr+).


Solid solution has also been rep~rted~~*"l for barium in the high-
temperature form of BaC12, most probably also possessing a fluorite type
of structure. For

Tm -T -- m
N
N-43
= ll00"C (observed e~perimentally~l) ,
N2(J) - N2(C) i l l 2 0.04
Lsm m
n=-
RT
x--
LJI2
N
4xlloo

2 X 1233
2 1.8 N" 2 .
Again, a solution mechanism corresponding either to Ba + Ba2+ + 2e-
and/or (less likely) Ba + Ba2+ + 2Ba+ seems indicated by these experi-
men-bal data. On the basis of activity measurements carried out by equi-
libration of MgC12 with MgAl alloys of known activity, Rogers, Todinson,
and Richardson45 suggest for Mg-MgC12 the solution mechanism Mg + Mg2+ +

nisms Me; + Mg2+ +,'e2 or Mg + Mg2+ -


( M ~ z ) ~ + . The precision attained does not exclude the alternate mecha-
2Mg+, and the data appear to fit
best a mixed mechanism, or in other words, an equilibrium 2e- + 2Mg2+
(Mg2)2+, or 2Mg+ e (Mg2)2f, discussed more generally below.
29

E l e c t r i c a l Conductance
W
Dworkin, Bronstein, and Bredig50,52’53 found it impossible t o apply
the method used f o r measuring conductivity i n t h e alkali-metal systems t o
t h e alkaline-earth systems. While the components, molten salt o r l i q u i d
metal, were found singly not t o react extensively, i f a t all, with t h e
synthetic sapphire c e l l , t h e i r mixtures d i d attack t h e l a t t e r readily.
Recourse w a s taken t o an all-molybdenum-metal apparatus embodying an as-
sembly of two p a r a l l e l rods as electrodes which were immersed i n t o t h e
melt from above t o a varying, accurately determined depth. With pure
molten salts the low resistance of t h i s c e l l l e d t o large e f f e c t s due t o
electrochemical polarization of t h e electrodes and resulted i n a large
frequency dependence of t h e measured resistance which could not be ex-
trapolated t o i n f i n i t e l y high frequency with any accuracy a t a l l . How-
ever, the addition of even small amounts of metal, acting as a depolar-
izer, removed t h e frequency dependence. Thus, it w a s possible t o use t h e
apparatus with t h e metal solutions and t o c a l i b r a t e it with solutions of
cadmium i n cadmium chloride. The specific conductivity of these solu-
t i o n s i s known from the early measurements of Aten i n 1910 (ref 5 4 ) and
was confirmed by new measurements with the sapphire cell.53
Only results f o r solutions of calcium and strontium i n t h e i r diha-
l i d e s , excepting t h e fluoride, have been reported thus far.46*50 It was
concluded, as far as the limited data permit, t h a t calcium behaves some-
w h a t l i k e sodium i n t h a t the rate of increase i n s p e c i f i c conductivity
,
w i t h m e t a l concentration decreases (Fig. 18). The same i s true f o r
: .
strontium i n SrBr2 a t the r e l a t i v e l y low temperature of 7OO0C, but i n the
dichloride, a t 910°C, t h e behavior of strontium, giving a l i n e a r increase
of conductance with metal concentration, l i e s somewhere between t h a t of
sodium and potassium, with barium expected t o resemble t h e l a t t e r . The
decreasing equivalent conductance of the calcium-metal solute was a t t r i -
buted t o t h e formation, not of diatomic molecules, C a 2 , as with sodium,
but of hypothetical molecule ions, (Ca2)2+, i n which t h e t r a p f o r elec-
trons i s t h e single two-electron bond similar t o that i n Na2, or i n Cd2*+
and Hg22+.
30

UNCLASSIFIED
ORNL- LR- DWG.70122

Fig. 18. Specific Conductivity of Alka-


line-Earth Metal Solutions in Alkaline-
Earth Halides (partially unpublished data).

4 .O
0
P I
I I
2
I
3
MOLE PER CENT METAL

From both the melting-point depression, above, and the decrease in


metal solute conductance from 800 to 450 ohm-’ an2 equiv-l, it would seem
to follow that at saturation (approximately 3 mole $) somewhat less than
half of the dissolved calcium metal is associated to Ca22+.
More information is certainly required to put these ideas about the
constitution of the solutions of the alkaline-earth metals in their di-
halides on a firm basis, and especially to exclude with certainty the
existence of I@ alkaline-earth ions in limited concentrations.

Rare-Eaxth-MetalA’rihalide Systems

The systems comprising rare-earth metals with their trihalides (see


Table 5 ) represent an especially interesting and more complex group of
metal-metal-halide systems, as they occupy an intermediate position.
They are farther than the alkaline-earth systems along the line leading
from the extreme of the alkali-metal solutions, with metallic properties
and with no solid subhalide fomtion, to the systems of the transition
. k i3

Table 5. Principal Data for Some R a r e - E a r t h - M e t a i d e Systems

1131 11.5 470 2.4 9.0 1099 None None None None 1 2.o 1273 1193 ll87
1061 (ll.8)(e) 375 (2.4) 14.5 1001 None None None None 15.5 ll73 ll93
1052 (12.0) 490 (2.8) 8.2 1007 10231 None None 1103c 33 ll.73 1193 ll73

1090 11.7 4.80 2.6 9.0 1050 None None None None 9 1173 1068
1005 (U.8) 345 (2.0) 12 .o 960 None None None None 14 1173 1068
1034 12.o 460 2.7 8.8 988 1004i None .None 10811 32 1173 1068 1064

1059 ll.4 565 3 .O 17.0 919 None 9321 None None 19 1073 1208
966 (12.0) 545 3.4 16 .O 852 8741 None None None 18 1023 1208
loll 12.6 500 3.1 11.9 939 946c None None 10311 29 1073 1208 1180

1032 ll.6 590 3.3 14.0 9W None 9533. 9751 11141 31 1173 1297
955 (12.0) 1297
1060 9.2 740 3.2 26.5 764 None None None 835C 37 1073 1297

1204 (12 ?) 555 (2.8 ?) 14.0 1098 None None None 11041 14 ll73 1585

1270 (10 ? ) 390 (1.5 ? ? ) 12.0 1221 None None None None 15 1423 1782

S. Eworkin and M. A. Bredig, J. Phys. Chem., March 1963.


(b)Liquidus slope at i n f i n i t e dilution, mostly from work by Corbett etal.
('In = (dT,/dNM)o (Sm/RTm).

(d)see r e f U.
(e)Figures i n parentheses are estimates.
32

and posttransition elements, where metal-salt m i s c i b i l i t y i n general be-


comes more limited, where e l e c t r i c a l conductivity of t h e solutions gives
no indication of metallic character, and where s o l i d "subhalides" do
occur.

Phase Diagram

~ ~t h e system Ce-CeCl3 in-


An e a r l y phase diagram by C u b i ~ c i o t t iof
dicating a s o l u b i l i t y i n cec13 as high as 33 mole $ cerium metal was
l a t e r 5 6 found t o be i n error: A s o l u b i l i t y value of 9 mole 46 cerium,
apparently r a t h e r constant between 780 and 88O'C, w a s obtained by Mellors
and S e n d e r ~ f f ,and
~ ~ confirmed by measurements of Bronstein, Dworkin, and
B ~ - e d i ga~t ~855'. On t h e metal side, t h e phase diagram proposed by
- Mellors and Senderoff seems t o require further investigation, as both t h e
t r u e melting point of cerium m e t a l (817OC) and a t r a n s i t i o n occurring i n
t h e s o l i d state (73OOC) were not taken i n t o account.
The phase diagram of t h e corresponding lanthanum system, as deter-
~ ~ 19), gives a s o l u b i l i t y of t h e
mined by Keneshea and C u b i ~ c i o t t i(Fig.
metal i n t h e salt very similar t o t h a t of cerium. No intermediate s o l i d
phases, t h a t is, no s o l i d compounds containing cerium or lanthanum i n a
valence state lower thm 3, were observed. I n this respect, then, these
two systems resemble the alkaline-earth systems. Also, i n both cases, it
appears from the melting-point depressions of t h e t r i c h l o r i d e s t h a t n,
t h e apparent number of new p a r t i c l e s produced on dissolving one atom of
metal, i s approximately 3 o r s l i g h t l y l e s s , perhaps 2.5.
This signifi-
cantly rules out t h e smaller values, n = 1 f o r the dissolution of these
= I
metals as atoms and n = 1.5 f o r
~ e ~ -,
1 3~ C ~ Co rI 2 ~ +
e Me~13 -M' ions58 formed according t o 2Ce
3 ~ e ( ~ e ~ l unless
complex ions were present i n pure CeC13, so t h a t (CeC14)'
4),
+
no o r few ( ~ e ~ 1 4 ) -
ions a l s o w o u l d
a c t as new particles, i n which case n would be approximately 3.No com-
pounds of monovalent rare-earth elements seem ever t o have been reported,
even though some of these elements have long been known i n t h e divalent
state, notably Yb2+, EU2+, and Sa2+, with 14, 7, and 6 electrons i n t h e
4f shell, t h a t is, with t h e 4f s h e l l f u l l , half-filled, o r nearly half-
f i l l e d respectively. We shall, however, see below t h a t an equation
4d
L
II - .
33
I

UNCLASSIFIED
O R N L - LR-DWG. 70745
400 I I I I I I I I

-
- 90
0

W
K
3
I-
a
K
W
n
3 801
0
THERMAL ANALYSIS
O

SALT- PHASE SAMPLED


s
0 QUENCHED

701 I I I I I I I I I 1
IO 20 30 40 50 60 70 80 90 (00
MOLE PER CENT L a

Fig. 19. Lanthanum Metal-Lanthanum Trichloride System [F. J.


Keneshea and D. Cubicciotti, I . Chern. Eng. Data 6, 507 (1961) (copy-
right 1961 by the American Chemical Society and reprinted by permission
of the copyright owner)].

Ce + 2Ce3+ - 3Ce2+ (cryoscopic n = 3) would not seem t o do j u s t i c e t o


t h e r e s u l t s of t h e measurements of e l e c t r i c a l conductance, which suggest
Ce 3 ce3+ + 3e-.
Extensive, careful work on rare-earth-metal-halide phase diagrams
i s being done by Corbett and a s ~ o c i a t e s . ~ ~Stable,
- ~ ~ s o l i d compounds
corresponding to, o r a t l e a s t approximating, t h e composition MX2 were
found t o e x i s t i n t h e systems Nd-NdC13, Nd-NdI3, Pr-PrI3, Ce-CeI3, and
La-LaI3, besides d i s t i n c t s o l i d compounds intermediate between MX2 and
MX3 such as NdC12.3
7
, NdC12,2 7 , P r I 2 . 5 , CeI2.4, and LaI2.4
2
(Figs. 20-23) .
The investigators appear t o have had reasons f o r accepting t h e latter,
nonstoichiometric compositions i n preference t o simple ones such as
NdCl302NdC12, NdC1303NdC12, and MT3*MI2 (M = Pr, Ce, or La). Some of
these compounds m e l t congruently, others incongruently. Significantly,
t h e s o l i d diiodides of lanthanum, cerium, and praseodymium a r e charac-
t e r i z e d by melting points higher than those of t h e t r i i o d i d e s (Figs. 22
and 23) and, above all, by high metal-like e l e c t r i c a l conductivity
x
(>1 lo3 ohm-’ cm-’), unique among halides, except perhaps f o r Ag2F.
Solid Nd11,95, on t h e other hand, has a d i f f e r e n t c r y s t a l structure, i s
!

34

UNCLASSI FIE0
ORNL-LA-DWG. 56015

g
w* t

Fig. 20. Neodymium Metal-Neodymium Trichloride System


(thermal analysis: +, equilibration: e) [L. F. Druding and J. D.
Corbett, J. Am. Chem. SOC. 83, 2462 (1961) (reprinted by per-
mission o f the copyright owner, the American Chemical Society)].

UNCLASSIFIED
ORNL- LR-DWG. 56816

Fig. 21. Neodymium Metal-Neo-


dymium Triiodide System (thermal
analysis: +, equilibration: 0 ) (L.F.
Druding and J. D. Corbett, J. Am.
Chem. SOC. 83, 2462 (196l)(reprinted
by permission of the copyright owner,
the American Chemical Society)].

MOLE%Nd IN NdIs G.
-..
UNCLASSIFIED
ORNL-LR-DW$. 70759
I I I I I I I IQ I I
880 - -
+ T H E R M A L ANALYSIS ::Tr
860 - 0 EQUILIBRATION l! 7
3
X LITERATURE I
840 - 0
*+-+-' -

- 820- +-
-
+

P
w 800-
(r
3

760-

I = 740 -

720 - L aI -
L0I2.q2 2.00
1
!
-
7000 d, A 4; tk Lb 214 '28 22' 26 "'916 (00

La13 MOLE PER CENT L a IN LaI, La

Fig. 22. Lanthanum Metal-Lanthanum Triiodide System [J.


D. Corbett et al.. Discussions Faraday SOC. 32, 81 (1961) (re-
printed by permission of the copyright owner, the Faraday
Society)].

UNCLASSIFIED
ORNL- L R - DWG. 70758
820 I I I I I I I / I I I '4
+ T H E R M A L ANALYSIS
0 EQUILIBRATION
780 X LITERATURE
++ +

640
I I I I I I I I I I "
0 4 8 42 16 20 24 28 32 36 40 700
PrI, MOLE PER CENT Pr I N Pr13 Pr

Fig. 23. Praseodymium Metal -Praseodymium Triiodide


System EJ. D. Corbett et al., Discussions Faraday Soc. 32, 81
(1961) (reprinted by permission of the copyright owner, the
u
2
Faraday Society)].
36

s a l t l i k e , nonconducting, and m e l t s 225' below N d I 3 (Fig. 21). The prop-


e r t i e s of the metal-like diiodides were interpreted as being due t o the
presence of "metallic" l a t t i c e electrons plus t r i p o s i t i v e m e t a l cations,
M3+e-(I")2, while t h e s a l t l i k e NdC12 i s simply Nd2+(C1-)2. This inter-
pretation i s f'urther supported by t h e r e l a t i v e weakness of t h e paramag-
netism of s o l i d La12 with a s u s c e p t i b i l i t y of 220 X emu/mole, in-
dicating t h e absence of La2+ ions.
The s o l u b i l i t y of t h e metal i n the t r i i o d i d e a t 830', t h a t is,
above t h e melting point of any diiodide, decreases from 37 mole
i n N d I 3 (Fig. 21) t o 29 i n &I3 (Fig. 23) and 30 i n Ce13, but rises again
t o 33 mole $ i n La13 (Fig. 22). A similar deviation from t h e trend i s
shown by lanthanum i n LaC13 with a s o l u b i l i t y of 10.3 mole $ metal a t
900' (Fig. 19), much higher than expected i n t h e series: 30 mole $ i n
NdC13 (Fig. 20), 20 i n PrC13,61*64 and 9 i n CeC13.53*56 The reason for
this "anomaly" with lanthanum i s not quite clear; it may, however, be
remembered t h a t La3+, l i k e Sc3+ and Ce4+, possesses, i n contrast t o the
other rare-earth ions, a noble-gas type of electronic s h e l l .
Examination of the liquidus curves for t h e metal-like diiodides
(Figs. 22 and 23) confirms t h e high degree of dissociation of molten
La12 i n t o La3+ ions: The depression of t h e melting point by t h e addition
of La13 (Fig. 22) i s only of t h e order of 0.50' per mole MX3, corre-
sponding, with an estimated heat of fusion of 10 kcal/mole, t o a cryo-
scopic n, t h a t is, number of p a r t i c l e s per t r i i o d i d e molecule dissolved,
of only 0.2. For Ce1260 these numbers are larger, namely, 1.16' per
mole $ C e 1 3 and n = 0.5, and are s t i l l s l i g h t l y l a r g e r for BIZ (Fig.
23): 1.4' per mole $ -I3 and n = 0.6, indicating l e s s dissociation ac-
cording t o M2+ --$ M3+ +
e-. The apparent alignment of C e I 2 with P r I 2
rather than with La12 again i s of i n t e r e s t and not fully explained.
The increase i n t h e s t a b i l i t y of the M2+ rare-earth ion with in-
creasing atomic number leads t o a sharp maximum with Eu2+, which contains
seven 4f electrons, t h a t is, half of t h e full 4f shell, and i s more
s t a b l e than Eu3+ even i n aqueous systems. T h i s s t a b i l i t y , which i s very
low with gadolinium, perhaps gradually rises again i n going toward t h e
well-established Yb2+, which contains t h e complete s e t of 14 4f electrons.

c
37

E l e c t r i c a l Conductivity

The f i r s t measurements of the e l e c t r i c a l conductance of solutions


of cerium i n cerium trichloride, by Mellors and S e n d e r ~ f f ,indicating
~~
a rapid r i s e i n conductivity up t o 0.65 mole $J metal followed suddenly by
a very slow rise above t h a t concentration, were shown by Bronstein,
Dworkin, and t o be i n e r r o r because of reaction of t h e solu-
t i o n s with ceramic crucible material. (The r e s u l t s of a study of emf and
t h e i r i n t e r p r e t a t i o n i n terms of a Ce+ ion65 must have been equally af-
fected.) Actually, t h e conductivity i n t h e Ce-CeC13 system, as i n La-
LaC13r50 rises with a monotonically increasing rate t o a value, i n t h e
saturated m e t a l solution, approximately f i v e times t h a t of t h e salt. In
t h e neodymium system t h i s r a t i o amounts t o only 1.7, and i n the praseo-
dymium system it i s intermediate, 2.5.66 A similar relationship holds
f o r t h e equivalent conductance of these solutions (Fig. 2 4 ) obtained with

UNCLASSIFIED
ORNL- LR -DWG. 70425

-E
-
‘E
r
0

Fig. 24. Specific Conductivity o Rare Earth >-


k
Metal-Rare Earth Metal Chloride Systems. 2

5 (0 45 20 25 30
MOLE PER CENT METAL
38

estimates of the equivalent volumes on t h e basis of t h e density measure-


ments of Mellors and Senderoff i n t h e CeCls-Ce system.58 The results
obtained with t h e corresponding iodide systems67 are not very different;
however, t h e increases f r o m t h e pure salts t o solutions containing, f o r ex-
ample, 10 mole $ metal, correspond t o f a c t o r s as high as 15, 19, 6, and 2
f o r the La, Ce, Pr, and Nd systems respectively. These findings may be ex-
plained i n terms of t h e systematics of t h e s t a b i l i t y of t h e divalent rare-
e a r t h cation, M2+, as observed i n t h e phase diagram studies above. Where
no s o l i d compound of the composition MX;! i s found, o r where such a s o l i d
exhibits metallic character, * 6o conductivity of t h e molten solution i s
high, indicating t h e equilibrium M2+ e M3+ + e' t o be far t o t h e right.
Praseodymium, forming no PrX2 i n the chloride system, but a s o l i d phase
P r C l z , 3 of mixed valency which i s s t a b l e i n a very l i m i t e d temperature
range,64 and forming, i n t h e iodide system, a diiodide, not s a l t l i k e
but metallic i n character,60 a l s o occupies an intermediate position be-
tween cerium and neodymium as far as conductivity i s concerned.

UNCLASSIFIED
ORNL-LR-DWG.70123
I

10
I
I
- 9 I
I
Fig. 25. Specific Conductivity in Rare
Earth Metal-Rare Earth Metal Iodide Systems
[R. A. Sallach et al., 1. Phys. Chern., in press
(1963) (reprinted by permission of the copyright
owner, the American Chemical Society)].

a
-.-

0 5 (0 15 20 25 30 35
I MOLE PER CENT METAL
39

bs O f special i n t e r e s t i s t h e maximum i n t h e curve of conductivity vs


concentration f o r solutions of neodymium i n N d I 3 a t approximately 23
mole $ neodymium, which corresponds t o a maximum deviation from addi-
t i v i t y of t h e conductivity of mixtures of N d I 3 and N d I 2 a t 16.7 mole '$
neodymium metal, corresponding t o 50 mole $ N d I z (Fig. 25). Bredig
e t a1.67 i n t e r p r e t t h i s maximum i n terms of an electron exchange between
Nd2+ and Nd3+ f o r which t h e probability has a maximum a t a concentration
r a t i o of N d I 2 z N d I 3 = 1:1, since both these salts as pure l i q u i d s conduct
only ionically. The same e f f e c t i s thought t o cause the curvature of
t h e conductivity-concentration curve i n t h e NdC13-NdC12 system, but,
probably because of the smaller p o l a r i z a b i l i t y of Cl- compared with I-,
the r a t e of t h e electron exchange i s lower and a maximum i s not produced.
I n any comparison of the conductance observed i n the rare-earth-
metal-halide systems, with both the alkaline-earth and alkali-metal
systems, it i s reasonable t o make t h e correlation f o r the equivalent
volume, t h a t is, a volume of solution containing t h e same r a t i o of t h e
number of obstacles affecting electronic conduction, namely, the number
of anions t o the number o f metal electrons. However, t h e number of
cations i n such a volume decreases i n going from t h e alkali-metal halides
t o the rare-earth-metal halides by a f a c t o r of 3 . Now, according t o
Bredig, 68 t h e electronic conductivity, t h a t i s , t h e contribution of the
metal i n t h e d i l u t e s t a t e , may be thought of as being proportional both
t o t h e t o t a l number of cations and t o t h e number of valence electrons
i.

I introduced by t h e addition of metal, t h a t is, t o the concentration of


t h e two components making up a metal. The product of these two quanti-
t i e s is, then, thought t o be a s i g n i f i c a n t parameter. It equals the
product of t h e equivalent o r mole f r a c t i o n of t h e metal, NM, representing
t h e concentration of electrons from M = Mz+ + ze', and t h e reciprocal of
t h e valency of M, l/z, representing t h e concentration of MZf ions.
Figures 26 and 27 show the contribution of the m e t a l solute per equivalent
of mixture , $*NM = Asoh - (1- NM )Asalt, p l o t t e d against NM/ z , f o r sev-
e r a l chloride systems. It i s c l e a r t h a t this simple correlation can be
expected t o be s i g n i f i c a n t only f o r those f e w MXZ-M systems i n which all
M atoms r e a c t according t o M 3 Mzs + ze-, or, i n other words, where there
i s no o r l i t t l e complication by the trapping of electrons i n p a i r s such
40
UNCLASSIFIED
ORNL-LR-DWG. 6 6 9 3 0

6
Fig. 26. Electronic Conductance in Some Metal-Metal
Chloride Systems. Left: Eq uivalent Conductance of Metal
vs Mole Fraction of Metal. Right: Electronic Contribution
to Equivalent Conductance of Solution vs Product of Elec-
tron and Cation Concentrations [M. A. Brkdig, J. Chem.
Phys. 37, 914 (1962) (reprinted by permission of the copy-
right owner, the American Institute of Physics)].

UNCLASSIFIED
ORNL-LR-DWG. 66931

6-

I
a04
I I
0.08
I I
0.42
,
0.46
V I I I I I
0.04 0.02 a03 a04

Fig. 27. Electronic Conductance i n Ce, La, and K


Metal-Metal Iodide Systems. Left: Equivalent Conductance
of Metal vs Mole Fraction of Metal. Right: Electronic Con-
L4
1
tribution to Equivalent Conductance of Solution vs Product of
Electron and Cation Concentrations (M. A. Bredig).
e
41

as i n molecules (e.g., Na2), i n molecule ions (Ca22+), o r singly i n lower


valence states such as Fr2+. Suitable systems appear t o be KC1-K, SrC12-
Sr, and La-Lac13 o r Ce-CeCl3. Figure 26 shows La-Lac13 and Ce-CeCl3 t o
have indeed a dependence of %*NM on NM/z rather similar t o t h a t of K-KC1.
It i s perhaps possible t o ascribe t h e f a c t t h a t the curve f o r Sr-SrC12 i s
considerably below the other ones t o the trapping of electrons i n (Sr2)2+
molecule ions. However, the attempt t o i n t e r p r e t these f i n e r d e t a i l s i s
not r e a l l y satisfactory, as i s i l l u s t r a t e d by Fig. 27, where $-NM for
CeI3-Ce first deviates negatively and t h e r e a f t e r positively from $-NM
f o r KI-K. An explanation does not seem available.
Similarly, t h e f a c t t h a t the r i s e i n % w i t h t h e metal concentration
i s much f a s t e r f o r cerium i n Ce13 (ref 69) than i n CeCl3 remains unex-
plained, except t h a t it possibly has something t o do with the s t a b i l i t y
of the m e t a l l i c a l l y conducting s o l i d CeI2 and t h e nonexistence of a
similar s o l i d CeC12.

"N0"bETAZ;LIC" METAL SOLUTIONS

The term used here t o c l a s s i f y a second l a r g e group of solutions of


metals, e s s e n t i a l l y t h e heavy metals, i n t h e i r molten halides i s based on
t h e observation, made on a number of such systems, t h a t t h e e l e c t r i c a l
conductivity of t h e melt i s not increased, i f a l t e r e d a t a l l , on addition
of m e t a l . I n contrast t o t h i s viewpoint about t h e nonmetallic character
of these solutions, Cubicciotti' has pointed t o the metallic l u s t e r which

. deeply colored (black) solutions of bismuth i n molten B i C 1 3 reportedly


exhibit. The significance of t h i s apparent discrepancy i s not clear, and
we s h a l l proceed with t h e description of t h i s second c l a s s of solutions,
bearing i n mind that f u r t h e r study may tend t o b l u r t h e sharp l i n e of
separation found t o be a convenience a t t h i s time.
There is, however, one more c r i t e r i o n , which r a t h e r sharply sepa-
r a t e s many of t h e systems t o be discussed below from t h e solutions of the
electropositive metals d e a l t with on t h e preceding pages: This i s t h e
formation of e l e c t r i c a l l y nonconducting, stoichiometric s o l i d halides of
the metal i n a low valence s t a t e . Corbett e t al. have suggested t h a t t h e
e, s o l u b i l i t y of t h e posttransition metals i n t h e i r molten normal halides
p a r a l l e l s or, rather, depends on t h e i r a b i l i t y t o form a cationic species
of lower than normal charge. We s h a l l see t h a t t h i s tendency increases
Table 6 . Solubility Of Some Transition and F a t t r a n s i t i o n Metals i n Their Molten Halides
~ ~~~~

Solubility Temp Solid Solubility Temp Solid


M-MXZ Refs M-MXZ Refs
(mole '$ M) ("C) "Subhalides" (mole '$ M) ("C) "Subhalides"

-
N i NiCl2 9.1 977 None 70 In-InC1 InC1, In31nC16
"1n~12"
97

( I N C h
Ag-AgC1 0.03 490 None 71 TI.-T1C1 0.09 550 TlCl n
0.06 700 0.09 650 (TlAlC14?)
m-mc4 0.18 500 None 80 Sn-SnC12 0.0032 500 71
1.64 600
zn-znI2 0.28 500 None 80 Sn-SnBrz 0.068 500 71
1.65 670
c a-C a c l 2 14.0 550 None 54,79 Pb-PbCl2 0.020 600 71
21.0 800 0 -052 700
30 .O 1000 0.123 800
C d- C dBr2 14.0 550 None 79 Pb-PbI2 0 0024 4-40
20 .o 700 0.15 600
28 .O 900 0.41 700
Cd-CdIz 2.5 400 None 79 Sb-SbCl3 0.018 270 80
15 .O 700
25 .O 950 ~b-sb13 1.69 200 a214 80, 102
3.5 300
Hg-HgC12 7.0 280 Hg2 c12 73 5.8 400
18.0 4.00
4.0 .O 500 Bi-BiC13 28.0 202 107
46 00 320 "BiCl"
Hg-Hgl2 25 .O 230 Hg2 I 2 76 28 .O
35 .o 280 100.o
Ga-GapCl4 3.7 180 n Bi-BiBr3 21.0 205 108
57.0 294 "BiBr"
Ga-GazBrq 14.0 180 95 45 .O 4-40
10000 >538
80
Bi-BiI3 48.0 336 108
10000 #t58 "MI"

4
43

rapidly within each group with increasing atomic weight (see Table 6 ) . It
a l s o depends on t h e halide ion, and Corbett e t al. a t t r i b u t e t h i s l a t t e r
trend t o t h e r e l a t i v e tendency toward s t a b i l i z a t i o n of t h e higher oxida-
t i o n s t a t e by complexing with the halide ion, normally i n t h e sequence
C1 > B r > I, but reversed f o r C d and Hg.

Transition Metals

The occurrence i n this group of elements of more than one s t a b l e


I valence s t a t e i s a w e l l established general feature of t h e i r chemistry.
i i
However, very l i t t l e seems t o be known about the i n t e r a c t i o n of these
metals with t h e i r molten salts. Solid salts of a monovalent s t a t e of the
m e t a l seem t o be almost e n t i r e l y missing, t h e bivalent s t a t e being t h e
lowest i n which t h e great majority of stable s o l i d o r l i q u i d halides of
the t r a n s i t i o n elements are known t o occur. There i s a notable exception
[besides t h e compound K N i ( C N ) 2 ] i n the study of the nickel-nickel chloride
systems by Johnson, Cubicciotti, and Kelly,7o i n which a s o l u b i l i t y as
high as 9 mole $ nickel metal was found a t t h e eutectic temperature of
980'C. The authors consider t h e formation of N i + ions from N i + Ni2+ +

2 N i + as being indicated by t h e freezing-point depression i n con junction


with t h e heat of fusion, but do not r u l e out t h e p o s s i b i l i t y of an alter-
nate, "physical, " interpretation.
However, i n t h e l a t t e r case an asymptotic approach of t h e a c t i v i t y
of N i C l 2 t o t h e curve of i d e a l a c t i v i t y vs concentration near = l
aNic1-2
would be expected but i s not found. A t h i r d p o s s i b i l i t y which was not
discussedby t h e authors, and t h e odds of which are hard t o evaluate i n
t h i s case, i s t h e s p l i t t i n g - o f f of two electrons, according t o N i --r Ni2+ +
2ea, similar t o t h e dissolution of electropositive metals i n t h e i r molten
halides discussed i n t h e preceding sections. E l e c t r i c a l conductance meas-
urements will have t o be made t o s e t t l e this question. Such measurements
are l i k e l y t o show e i t h e r very l i t t l e change i n conductance, indicating
N i + ions, o r some s m a l l rise resembling t h a t i n t h e NdC13-NdC12 system,
above, e n t a i l i n g a similar i n t e r p r e t a t i o n i n terms of electron exchange
between metal ions of different valency, N i + and Ni2+. A large rise i n
W conductivity, corresponding t o the introduction of t r u l y mobile electrons,

.
t h a t is, of electrons i n shallow, F-center-like traps, i s considered t h e
t h i r d , least l i k e l y p o s s i b i l i t y .

Posttransition Metals

It appears t h a t i n t h e first B subgroup of t h e periodic system, only


the s o l u b i l i t y of s i l v e r i n s i l v e r chloride has been determined and found
by Corbett and Winbush71 t o be as low as 0.06 mole $ a t 7OOOC. The only
s o l i d subhalide known i n this group i s Ag2F, prepared i n aqueous solution
and readily decomposed a t temperatures above 90°C. 72
I n t h e second B subgroup the products of i n t e r a c t i o n of mercury metal
with i t s d i h d i d e s , t h a t is, t h e "subhalides" Hg2X2 of mercury, have been
w e l l known for a long time and are stable, even i n contact with water.
The phase diagram of t h e (anhydrous) HgClp-Hg system (Fig. 28) as deter-
mined by Yosim and M a ~ e shows
r ~ ~ s o l i d Hg2Cl2 t o be stable up t o the
"syntectic" temperature of 525O, where it decomposes i n t o a s a l t - r i c h melt
of almost t h e same composition and a s m a l l amount of a l i q u i d mercury
phase containing 6.8 mole $I HgC12 i n solution. With t h e use of a calori-
metric value of the heat of f'usion of HgC12,74 t h e depression of i t s

UNCLASSIFIED
ORNL-LR-DWG. 56700
6001
600 I , 1 I I
I

: 5+L2
550 - I
I

RUFF 8 SCHMOER
SMITH 8 MENZIES

THIS WORK
QTHERMAL LJATA
AVISUAL DAm

MOLE PERCENT Hg

Fig. 28. Mercury-Mercuric Chloride System [S.


J. Yosim and S. W. Mayer, I. Pbys. Chem. 64, 909
(1960) (reprinted by permission of the copyright
owner, the American Chemical Society)].
CI
45

freezing point on addition of mercury metal was interpreted i n terms of


t h e dissolution e i t h e r as mercury atoms o r as undissociated Hg2C12 mole-
cules from Hg + HgC12 * HgzCl2. According t o a thermodynamic analysis
of t h e liquidus curve f o r Hg2C12 by Yosim and Mayer,73 Hg2C12 i s a more
l i k e l y solute species than -the mercury atom i n that concentration region.
Since Hg2C12 does not impart e l e c t r i c a l conductivity t o i t s solution i n
molten HgC12,75 the former must be undissociated, as i s the latter. The
deepening i n color of these solutions e i t h e r with increasing temperature
o r mercury-metal content merits f u r t h e r investigation.
A study of t h e HgI2-Hg system by Pelabon and Laude, 1929,76 seems
t o be too incomplete t o allow d e f i n i t e conclusions t o be drawn.
The cadmium halide systems have been77 and s t i l l a r e under quite
extensive investigation. With one exception, 78 most probably i n error,
i f f o r no other reason but t h a t the reported phase diagram f o r CdC12-Cd
v i o l a t e s the phase rule, no s o l i d subhalides Cd2X2 have been observed.
However, considerable s o l u b i l i t y of metal i n t h e molten halides e x i s t s ,
and it r i s e s with temperature.79 Like t h a t i n the alkali-metal halide
systems, t h e s o l u b i l i t y seems t o be l a r g e s t i n t h e bromide and lowest i n
t h e iodide system. A very low value of 1.5 mole a t 600' given f o r the
iodide system,80 however, seems t o be i n error, compared with 10 mole $,
according t o Topol and L a n d i ~ . ~ ' The l a t t e r determination i s not com-
p l e t e l y s a t i s f a c t o r y either, because of a probably too rapid r i s e i n
s o l u b i l i t y a t low temperature which produces an e n t i r e l y improbable
curvature i n a logarithmic p l o t of s o l u b i l i t y vs 1/~. Consolute t e m p e r a -
tures may be estimated t o be very high, i n t h e general v i c i n i t y of 1500',
and a r e not v e r i f i a b l e because of high vapor pressure.
A considerable amount of data e x i s t s f o r t h e cadmiwadmium chloride
system.77s81-83 AtenS4 proved t h e solutions t o be not c o l l o i d a l i n nature
because of the l a r g e melting-point depression of CdC12 on addition of cad-
mium metal. He a l s o showed t h a t t h e e l e c t r i c a l conductivity does not in-
crease, but r a t h e r decreases s l i g h t l y . Farquarson and H e ~ m a n nfound
~~
t h e solutions t o be diamagnetic, and t h i s w a s confirmed by Grjotheim
-
e t al.85 and by Nachtrieb.86 It indicated that the metal dissolves
e i t h e r i n t h e form of atoms, molecules, o r molecule ions such as (Cd2)2+,
analogous t o ( H ~ z ) ~ ' , and not as Cd+ i o n s O B 7 Cryoscopic data85'79 to-
gether with a recent calorimetric heat of of 7.22 kcal/mole f o r
i

4.6
i
C d C l 2 lead t o t h e suggestion of i d e a l behavior f o r concentrations of
Cd2C12 i n CdC12 below approximately 2 mole $ Cd2C12, but increasing nega-
t i v e deviation from Raoult's l a w f o r CdC12 with increasing metal (Cd2Cl2)
solute concentration. This negative deviation (7 CdC12 = 0.96 o r 0.99
f o r a cadmium atom o r a C d 2 C l 2 solute, respectively) a t t h e monotectic
(15 mole C d 2 C 1 2 ) may be explained by the substitution of large cations
(Cd2)2+ f o r smaller Cd2+ ions. It i s analogous t o similar deviations
found i n systems with other l a r g e cations such as KCl-CdC12 and may or
may not be thought of i n terms of t h e formation of complex anions such as
(CdC14)2'.88 Perhaps t o explain t h e asymptotic decrease i n the negative
deviation with decreasing metal. concentration, t h e formation of undis- -
sociated species Cd4C16 and Cd3X4 has been suggested,79 but does not ap-
pear plausible i n a highly ionized and ionizing medium such as molten
CdC12. An alternative explanation i s t h a t i n s m a l l concentrations t h e
complex anion m a y a c t as a "common ion" w i t h respect t o some of the ions
i
Of t h e solvent salt, CdC12, or, i n other words, t h a t a c e r t a i n concentra-
t i o n of complex anions such as (CdC14)2' i s present, even i n pure molten
CdC12. Solid solution f o m t i o n which would invalidate any such i n t e r -
pretations of cryoscopic data i s deemed unlikely.
The observed negative deviation from Raoult's l a w would be more dif-
f i c u l t t o reconcile with t h e "physical" model (cadmium atoms), although
Nachtrieb86 points out that t h e t r u e s t a t e of affairs may be a subtle one,
perhaps corresponding t o a cadmium atom "solvated" by a Cd2+ ion82 o r t o
a solute species Cd22+ with an asymmetric charge distribution.
Observations by Grjotheim e t and by Hertzog and on t h e
e l e c t r i c a l transport of Cd with respect t o C1- were interpreted i n terms
of strong interaction between Cd and Cd2+. Other e f f e c t s such as those
produced by t h e addition of a t h i r d component upon the s o l u b i l i t y of
cadmium metal may be taken as an indication t h a t a fairly stable (Cd2)2+
ion exists. According t o Cubicciottigo t h e addition of KC1 t o CdC12 grad-
-
u a l l y lowers t h e s o l u b i l i t y of cadmium i n the m e l t . Corbett e t al.
explained this by assuming t h a t t h e excess chloride ions s t a b i l i z e the
higher oxidation s t a t e of cadmium by complexing. 80*88s 91 However, by
adding A l C l 3 (a strong Lewis acid) they prepared a solid, diamagnetic
compound of t h e formula Cd2(A1Cl4), containing a l l of t h e cadmium i n oxi-
dation s t a t e I.92 I n the melts t h e increased s t a b i l i t y of t h e cadmium(1)
47
i

oxidation state when the halide ion X 7 (Cl-, Br-, I-) i s replaced by the
l a r g e r AlX4' w a s considered t o r e s u l t from the decrease i n t h e i n t e r a c t i o n
of t h e Cd2+ cation with t h e anion, and, i n t h e solid, from t h e decrease i n
t h e difference i n l a t t i c e energies of t h e salts i n t h e two oxidation
states. The contrast between t h e dark-brown color of melts containing no
A l C l 3 and t h e light-green color of those i n which A l C l 4 - i s present was
a t t r i b u t e d t o the role, i n t h e former case, of t h e (highly polarizable!)
uncomplexed halide ion as a bridge i n a weak association between t h e ions
i n t h e two d i f f e r e n t oxidation s t a t e s , Cd22+ and Cd2+. Other examples of
deep color a r e found even i n aqueous systems containing Fe, Sn, Sb, o r Cu
i n two valence states.93
. I n the zinc systems t h e s o l u b i l i t y of t h e metal does not much exceed
1 mole $.80 Because of t h e strong halide complexing tendency of !Zn2+, it
i s doubtrul whether addition of A l X 3 would r a i s e the s o l u b i l i t y of zinc
i n znX2 t o any considerable extent by allowing t h e lower valence s t a t e
t o+form
(~n~~ ? ) i n greater concentrations.
I n the t h i r d B subgroup of the periodic system t h e halides of mono-
valent indium and thallium have long been known, and some of those of gal-
l i u m have a l s o been prepared recently.9698 "Dihalides" of these elements
have been recognized as ionic salts of t h e monovalent metal with anionic
complexes of t h e t r i v a l e n t metal, f o r example, Ga(I)[Ga(III)C14] .99-101
-
Again, A l C 1 3 was employed by Corbett e t a1.91994995998 t o a i d i n the cam-
p l e t i o n of t h e reaction 2Ga -t Ga3+ 3 3Ga+, t h a t is, t o convert a l l of t h e
gallium i n t o t h e low oxidation state (I), as i n Ga(I)AlC14.91 The solu-
! b i l i t y of aluminum m e t a l i n molten A l 1 3 w a s reported71 t o be as high as
0.3 mole $ a t 423OC.
I n t h e fourth group of posttransition elements, dihalides of germa-
nium as w e l l as those of t i n and l e a d a r e w e l l known stable compounds.
The s o l u b i l i t i e s of the l a t t e r metals i n t h e molten dihalides are very
small, 719 go and it does not seem t o have been shown as yet t h a t addition
of Al.C13 increases t h e solubility, although this would seem t o be a t e s t
f o r t h e existence of an oxidation s t a t e lower than 11.
A l l t h r e e elements of t h e f i f t h posttransition group, arsenic, anti-
mony, and bismuth, possess a very s t a b l e oxidation s t a t e (111). The
s o l u b i l i t y of t h e m e t a l i n t h e molten t r i h a l i d e s r i s e s rapidly with in-
!
! - creasing atomic number, and i n t h e bismuth systems even leads t o t h e for-
48

mation of s o l i d "subhalides" of t h e approximate stoichiometry BiX. Vapor


pressure and emf studies of t h e %I3-%system by Corbett e t - al.1029103

demonstrated t h e existence of a catenated 12%-Sb12 species. Older


claimslo4 f o r t h e existence of dihalides BiX2 of bismuth were not substan-
t i a t e d later,1o5 especially not i n more recent phase behavior and r e l a t e d
studies by Sokolova, Urazov, and Kuznetsovlo6 and p a r t i c u l a r l y by Yosim
-
e t al.107-10 (Figs. 2S31). The l a t t e r study i s distinguished by t h e ex-
perimental feat of determining t h e consolute temperature (778") under a
BiC13 pressure of 80 atm. The l a t e s t suggestion f o r s o l i d bismuth "mono-
chloride," "BiC1," on the b a s i s of x-ray diffraction measurements, 111 is

the stoichiometric formula B i & 1 7 or ( B i 9 ) ( B I C 1 5 ) 4 ( B i 2 C l s ) , allegedly con-


taining, besides anions BiC132- and Bi2C1s2-, t h e remarkable complex
cation B i g 5 + corresponding t o t h e unique oxidation s t a t e of 5/9. On t h e
other hand, i n t h e s o l i d complex compound with A1Cl3, BiAlC14 (first pre-
-
pared by Corbett and McWlangl), Levy e t a l . l 1 2 propose t h e trimeric

Fig. 29. Bismuth Metal-Bismuth


Trichloride System [S. J. Yosim et
al., 1. Phys. Chem. 63, 230 (1959)
(reprinted by permission of the copy-
right owner, the American Chemical
Society)] .

5 DECANTATION P
li~
! 49
! i

; u
!
form of oxidation s t a t e (I), namely, triangular (Bi3)3+, which they a l s o
f i n d i n the molten s t a t e t o represent the b e s t interpretation of radial
i
distribution f'unctions obtained by x-ray diffraction.
Various other studies have been made on t h e molten solutions of
bismuth m e t a l i n i t s t r i h a l i d e s Keneshea and Cubicciotti have meas-
ured t h e vapor pressures of BiX3 over these solutions1l6l6 and t h e i r

500

0
400
a
3

igf 300
E!
200

o-BlBr3+ BI Br

400

300

200
I I I
i a
a

100

W
! -
50

densities. 7*l1 7s While the chloride and bromide systems show large
positive deviation from Raoult's law for the solvent, BiX3, when atoms
are assumed to be the solute species, Bi13 seems to behave ideally up to
30 mole $ bismuth. The reasons for this peculiar behavior of the iodide
system are not apparent. Bredigflg interpreted the vapor pressure meas-
urements in the Bi-BiBr3 system in terms of an equilibrium between Biz
molecules, ions, and BiBr3, namely, 2Bi2 + ai3+ -+ 3(Bi~)~+,but
an equilibrium Biz + Bi3+ might explain these data equally
At higher temperatures the equilibrium appears to shift to Biz
and its dissociation products, Bi3+ and electrons. A multiple mechanism
for the dissolution of bismuth in BiC13 was also proposed by Mayer, Yosim,
and T0p01.109*120 Corbett121 suggested an interpretation of the same
vapor pressure data in terms of a tetramer, (BiC1)4, possibly with Bi-Bi
bonds to explain the diamagnetism of solid 1tBiC1."122 The diamagnetism
of the molten solutions was discussed by Nachtrieb.86 On the basis of emf
and polarographic measurements, Topol, Yosim, and O ~ t e r y o u n g pro-
~~~*~~~
pose the existence of an equilibrium nBiX eBinXn in both the chloride and
bromide systems, with n = 4 apparently given preference. An equilibrium
of this nature is also indicated by the optical absorption measurements
of Boston and Smith,125 in which the ratio of two optically distinct
species, probably the monomer and a polymer of Bi+, was shown to depend
on the bismuth-metal concentration. Thus, the majority of authors are
found to prefer the assumption of polymeric subhalide formation in solu-
tions of bismuth in the trihalides to that of bismuth atoms.7*117*118
The early measurements of the electrical conductance by Aten126*127 showed
a decrease in the equivalent conductivity of solutions with metal concen-
tration. This is much better understood in terms of subhalide and Biz
molecule formation than of dissolution of bismuth atoms.
The solubility of salts of several posttransition metals in the
liquid metals was dealt with by Yosim and Luchsinger.128 At sufficiently
high temperatures bismuth halides are completely miscible with bismuth
metal; Hg2C12 was found to be soluble in mercury metal to the extent of
7 mole $ at 60OOC. The solubility of PbC12 in lead at 1000° is given as
1 mole $, while the salts of the remaining metals are considered insoluble
in their metals. There appears to be no simple relationship governing
this behavior.
The question of specificity was also examined, that is, the question
of whether metals dissolve s a l t s other than their own halides. The solu-
51

b i l i t y of a foreign salt can be explained i n terms of oxidation-reduction


reactions t o form the halide of t h e solvent metal, which then dissolves
i n t h e metal.

SUMMARY

Mixtures of metals with t h e i r molten halides a r e not c o l l o i d a l sus-


pensions, but t r u e solutions. On t h e basis of both thermodynamic and
e l e c t r i c a l conductance measurements, it i s c l e a r t h a t many of t h e mixtures
are solutions i n which t h e electrons introduced by t h e metal, especially
n
i n t h e case of electropositive metals, are i n shallow t r a p s and conse-
,L. quently mobile, as, f o r example, i n potassium o r i n lanthanum systems.
I
The mixtures a l s o include various other types of solutions i n which t h e
electrons a r e p a r t l y o r wholly i n t r a p s of greater depth. Such t r a p s a r e
believed t o be diatomic metal molecules such as Na2; diatomic molecule
ions such as Hg22Sr Cd2'+, Ca22+; more complex cations such as Bi33+,
Bib4+, o r perhaps even Big5+; and simple monomeric ions of t h e metal. i n a
lower than "normal" valence state such as B i + and Nd2+. Electron exchange
between cations of different valence s t a t e s of t h e metal seems t o occur
and t o contribute s l i g h t electron mobility. The electronic conduction
process and t h e state of the electron i n solutions, such as the alkali-
metal solutions,in which the electrons a r e quite mobile, a r e not too well
understood as yet. However, there a r e a f e w first attempts t o describe
t h e o r e t i c a l l y t h e s t r u c t u r e and e l e c t r i c a l behavior of such solutions i n
terms of F (color )-center-like electrons, but t h e degree of delocalization
! *
of t h e mobile electron or, i n other words, t h e number of metal cations
with respect t o which each electron may be considered quantized needs
further c l a r i f i c a t i o n . Solute m e t a l atoms resembling gaseous atoms do
not deserve serious consideration, as t h e i r very high p o l a r i z a b i l i t y must
l e a d t o strong i n t e r a c t i o n with the ions of the molten salt, t h a t is, for-
mation of F-center-like configurations, i f not w i t h t h e molecule ions o r
similarly deep t r a p s j u s t mentioned above. Measurements of magnetic sus-
c e p t i b i l i t y , of paramagnetic and nuclear magnetic resonance, and of t h e
H a l l effect may perhaps bring f u r t h e r enlightenment regarding t h i s more

bd detailed aspect of the shallow t r a p s i n which t h e electrons contributing


' A considerable electronic conductivity a r e thought t o be located.
52

REFERENCES

1. D. Cubicciotti, USAEC Document MDDC-1058 (1946).


2. P. Drossbach, Elektrochemie Geschmolzener Salze, Springer, Berlin,
1938.
3. R. Lorenz and W. Eitel, Akadem. Verl. Ges., Leipzig, 1926.
4. W. Nermst, Theoretische Chemie, 8.-10. ed., p 468, Stuttgart, 1921.
5 . D. Cubicciotti, J. Chem. Educ. -337 540 (1960).
-
6. D. Cubicciotti, Transactions Aim, TP3589D, J. Metals 1953, 1106.
7. F. J. Keneshea, Jr., and D. Cubicciotti, J. Phys. Chem.62, 842-47
.
(1958)
8. R. B. Ellis, Chem. Eng. News 38 96 (1960).
9. E. A. Ukshe and N. G. Bukun, Russ. Chem. Rev. (English Transl.) -930
90 (1961).
.
10 E. D. Eastman, D. D. Cubicciotti, and C. D. Thurmond, "Temperature-
Composition Diagrams of Metal-Metal H a l i d e Systems'' in The Chemistry
and Metallurgy of Miscellaneous Materials. Thermodynamicst ed. by
L. L. Quill, National Nuclear Energy Series IV-l9B, p 6, McGraw-
Hill, 1950.
ll. M. A. Bredig, J. W. Johnson, and W. T. Smith, Jr., J. Am. Chem.
-
SOC. 77, 307 (1955).
12 Metals Handbook, vol. 1, 8th Ed., p 25, American Society for Metals,
0

1961.
-
13. H. REgener, Ann. Physik 29, 386 (1937).
14 A. S. DworMn H. R. Bronstein, and M. A. Bredig, J. Phys. Chem.
-
66, 572 (1962).
-
15. M. A. Bredig and H. R. Bronstein, J. Phys. Chem. 64, 64 (1960).
-
16. J. W. Johnson and M. A. Bredig, J. Phys. Chem. 62, 604 (1958). i
t
17 M. A. Bredig, H. R. Bronstein, and W. T . Smith, Jr., J. Am. Chem.
-
SOC. 77, 1454 (1955).
18. Iu. K. Delimarskii and B. F. Markov, Electrochemistry of Fused
-
Salts (English translation), Sigma Press, Washington, D.C., 1961.
19. M. A. Bredig and J. W. Johnson, J. Phys. Chem. -
64, 1899 (1960).
20. K. Pitzer, "The Nature of Chemical Bonding in Fused Alkali Halide-
Metal Systems," 18th International Congress of Pure and Applied
Chemistry, Montreal, Canada, Aug. 6-12, 1961.
21. K. Pitzer, J. Am. Chem. SOC. -984 202!5-28 (1962).
.
22 Milton Blander, private communication.
.-
23 D. R. Stuff and G. L. Sinke, Thermodynamic Properties of the Ele-
ments, American Chemical Society Monograph, 1956.
53

.
24. M. Blander (ed.), Selected Topics i n t h e Physical Chemistry of
Molten S a l t s ( i n preparation)
25. E. Mollwo, Chem. Abstracts 30, 47574 (1936).
26. -
E. Mollwo, Physik. Z. 36, 74.0 (1935).
27.
(1958).
-
H. R. Bronstein and M. A. Bredig, J. Am. Chem. SOC. 80, 2077

28. -
H. R. Bronstein and M. A. Bredig, J. Phys. Chem. 65, 1220 (1961).
29. S. A. Rice, Discussions Faraday SOC. -332 181 (1961).

30. Judith Brown, UCRL-9944 (Dec. 13, 1961).


31. A. Smakula, Nachr. Ges. Wiss. a t t i n g e n , I1 -
1(4),5 5 6 5 (1934).
32. H. R. Bronstein, A. S. Dworkin, and M. A. Bredig, J. Chem. Phys.
i -
34, 1843-4.4 (1961).
33. H. R. Bronstein, A. S. Dworkin, and M. A. Bredig, J. Chem. Phys.
-
37, 677-78 (1962).
-
34. L. Wzhler and G. Rodewald, Z. Anorg. Allgem. Chem. 61, 54 (1909).
I

.
35. D. D. Cubicciotti and C. D. Thurmond, J. Am. Chem. SOC. 71, 2149
(1949)
-
36. M. A. Bredig and J. W. Johnson, unpublished data, 1951.
-
37. P. Ehrlich and L. Gentsch, Naturwissenschaf'ten 4.0, 460 (1953).
-
38. P. Ehrlich and L. Gentsch, Naturwissenschaften 41, 211-12 (1954).
1

58-73 (1956) .
39. P. Ehrlich, B. A l t , and L. Gentsch, Z. Anorg. Allgem. Chem. -9283

-
40. H. Schzfer and A. Niklas, Angew. Chem. 64, 6l.l (1952).
41. J. A. Hinkebein, Ph.D. Thesis, Iowa S t a t e College, 1958.
-,
42. D. Peterson and J. A. mnkebein, J. Phys. Chem. 63 l.36-3 (1959).
43. M. A. Bredig, Proceedings, High-Temperature Symposium, Berkeley,
C a l i f . , 1956.
4.4. L. I. Staffmsson, "The Physical Chemistry of Metals i n Their
Molten Halides," Ph.D. Thesis, London, December 1959.
45. P. S. Rogers, J. W. Tomlinson, and I?. D. Richardson, p 909 i n
International Symposium on t h e Physical Chemistry of Process Metal-
-
lurgy, Pittsburgh, 1959, Ed. by G. R. St. Pierre, Interscience,
New York, 1961.
46. A. S. Dworkin, H. R. Bronstein, and M. A. Bredig, J. Phys. Chem.
( t o be published).
! 47. A. I. Burin, Metallurg -10(4), 87 (1935).
48. B. F. Naylor, J. Am. Chem. SOC. 150 (1945).
W
A
54

-
49. E. Mollwo, Nachr. Ges. Wiss. G%tingen, I1 1(6), 79-89 (1934).
50. A. S. Dworkin, H. R. Bronstein, and M. A. Bredig, Discussions
-
Faraday SOC. 32, 188 (1962).
-
51. A. S. Dworkin and M. A. Bredig, J. Phys. Chem. 67, 697 (1963).
52. H. R. Bronstein, A. S. Dworkin, and M. A. Bredig, J. phys. Chem.
-
64, 1344 (1960) .
53. H. R. Bronstein, A. S. Dworkin, and M. A. Bredig, J. Phys. Chem.
-,
66 44 (1962).
-
54. A. H. W. Aten, Z. Physik. Chem. 73, 578-97 (1910).
-
55. D. Cubicciotti, J. Am. Chem. SOC. 71, All9 (1949).
-
56. G. W. Mellors and S. Senderoff, J. Phys. Chem. 63, 1111 (1959).
57. F. J. Keneshea, Jr., and D. Cubicciotti, J. Chem. Ehg. Data 5,
507-9 (1961) b

-
58. G. W. Mellors and S. Senderoff, J. Phys. Chem. 64, 294 (1960).
59. 3. D. Corbett, F. Druding, and C. B. Lindahl, J. Inorg. Nucl.
Chem. -'
17 176-77 (1961).
-
60. J. D. Corbett e t al., Discussions Faraday SOC. 32, 79-83 (1961).
61. L. F. Druding and J. D. Corbett, American Chemical Society Mtg.,
New York, September, 1960. Abstracts page 2 U (No. 50).
-
62. L. F. Druding and J. D. Corbett, J. Am. Chem. SOC. 81, 5512 (1959).
-
63. L. F. Druding and J. D. Corbett, J. Am. Chem. SOC. 83, 246247
(1961). '
64. J. D. Corbett -ee t al 9 J. Am. Chem. Soc. ( t o be published).
65. S. Senderoff and G. W. Mellors, J. Electrochem. ->105 224 (1958).
66. A. S. Dworkin, H. R. Bronstein, and M. A. Bredig, J. Phys. Chem.
66, 1201-2 (1962).
67. M. A. Bredig e t al., "The E l e c t r i c a l Conductivity of Solutions of
Metals i n Their Molten Halides. V I . Lanthanum, Cerium, Praseo-
dymium, and Neodymium i n Their Molten Iodides," J. Phys. Chem. (to
be published i n May or June 1963).
-
68. M. A. Bredig, J. Chem. Phys. 37, 91615 (1962).
69. M. A. Bredig -*,e t al "The E l e c t r i c a l Conductivity of Solutions of
Metals i n Their Molten Halides. V I . Lanthanum, Cerium, Eraseo-
m u
m , and Neodymium i n Their Molten Iodides," J. Phys. Chem. ( t o
be published i n May or June 1963).
70. J. W. Johnson, D. Cubicciotti, and C. M. Kelley, J. Phys. Chem. -362
1107-9 (1958).
-
71. J. D. Corbett and S. V. Winbush, J. Am. Chem. SOC. 77, 396446
(1955)
-
72. A. Hettich, Z. Anorg. Allgem. Chem. 167, 67 (1927).
5
55

, w 73 . -
S. J. Yosim and S. W. Mayer, J. Phys. Chem. 64, 909 (1960).
74 . -
L. E. Topol and L. D. Ransom, J. Phys. Chem. 64, 1339 (1960) .
75 . H. Foote and N. Martin, Am. Chem. J. 4l, 451 (1908).
76
. -
H. Pglabon and Ifme. Laude, B u l l . SOC. Chim. France s e r 4 45, 488-92
(1929)
77.
(1930) .
G. V. Hevesy and E. Lijwenstein, Z. Anorg. Allgem. Chem. -3187 266-72

78. G. G. Urazov and A. S. Karnaukov, Dokl. Akad. Nauk


(1954) . -
SSSR 96, 535

79 . -
L. E. Topol and A. L. Laudis, J. Am. Chem. SOC. 82, 6291-93 (1960).
rr( 80 .-
J. D. Corbett,
79, 3020-24 (1957).
S. V. Winbush, and F. C. f i b e r s , J. Am. Chem. SOC.

i 81 . -
R. E. Hedger and H. Terrey, Trans. Faraday SOC. 32, 1614 (1936).
82 . -
(1930).
R. E. Heymann and E. Friedlgnder, Z. Physik. Chem. A148, 177-94

83. E. Heymann, J. Martin, and M. Mulcaby, J. Phys. Chem. -


47, 473-84
(1943).
84 . J. Farquarson and E. Heymann, Trans. Faraday SOC - .
31, 1004 (1935).
85 . K. Grjotheim, F. Grzonvold, and J. Krogh-Moe,
5824 (1955).
J. Am. Chem. SOC. 77-,
86 . -
N. H. Nachtrieb, J. Phys. Chem. 66, ll63-67 (1962).
87 -
J. D. Corbett, Inorg. Chem. 1, 700 (1962).
88 . -
M. A. Bredig, J. Chem. Phys. 37, 451-52 (1962).
89 . W. Hertzog and A. Klemm, Z. Naturforsch. -
15a, llOO-llO7 (1961).
90 . -
D. Cubicciotti, J. Am. Chem. SOC. 74, l l 9 8 (1952).

.
91. J. D. Corbett and R. K. McIvfullan, J. Am. Chem. SOC. 78, 2906-7
(1956)
-
92 J. D. Corbett, W. J. Burbard, and L. F. Druding, J. Am. Chem. SOC.
83, 76-80 (1961).
93. N. Davidson e t al -
J. Am. Chem. SOC. 73, 1946 (1951).
94. -
J. D. Corbett and R. K. McMullan, J. Am. Chem. SOC. 77, 4217 (1955).
95 . -
J. D. Corbett and A. Hershaft, J. Am. Chem. SOC. 80, 1530 (1958).
96 R. Carlston, E. Griswold, and J. Kleinberg, J. Am. Chem. SOC. 80 -,
1532-35 (1958)
97 R. J. Clark, E. Griswold, and J. Kleinberg, J. Am. Chem. SOC. -980
4764 (1958).
98 . -
R. K. McMullan and J. D. Corbett, J. Am. Chem. SOC. $0, 4'761-64
(1958).
99. L. A. Woodward, G. Carton, and H. L. Roberts, J. Chem. SOC. -1956
)

3721-22
56

100. L. A. Woodward, G. Garton, and H. L. Roberts, J. Chem. SOC. 1956 -,


3723 .
-
101. G. Garton and H. M. Powell, J. Inorg. Nucl. Chem. 4, 84 (1957).
-
102.
.
B. L. Bruner and J. D. Corbett, J. Inorg. Nucl. Chem. 20, 62-65
(1961)
103. -
J. D. Corbett and F. C. Albers, J. Am. Chem. SOC. 82, 533 (1960).
104. W. Herz and A. Guttmann, Z. Anorg. Allgem. Chem. -
56, 422-32 (1908).
105. -
B. Eggink, Z. Physik. Chem. 64, &9 (1908).
. 106. M. A. Sokolova, G. G. Urazov, and V. G. Kurnetsov, Khim. Redkikh
1,
(1954) .
Elementov,Akad. NaukSSSR,Inst. Obshch. i Neorgan. Khim. 102-14

107. S. J. Yosim e t al -
J. Phys. Chem. 63, 230 (1959).
108. S. J. Yosim e t al., J. Phys. Chem. -
66, 28-31 (1962).
64,
109.
.
S. W. Mayer, S. J. Yosim, and L. E. Topol, J. Phys. Chem.
2 3 W (1960)
llo. -
A. J. D a r n e l l and S. J. Yosim, J. Phys. Chem. 63, 18-15 (1959).
ill. A. Hershaft and J. D. Corbett, J. Chem. Phys. -
36, 551-52 (1962).
112. -
H. A. Levy e t al., -
J. Phys. Chem. 64, 1959 (1960).
-
113. G. Cleary and D. Cubicciotti, J. Am. Chem. SOC. 74, 557 (1952).
ll4. D. Cubicciotti, F. J. Keneshea, and C. M. Kelley, J. Phys. Chem.
-
62, 463-67 (1958).
115. D. Cubicciotti and F. J. Keneshea, Jr., -
J. Phys. Chem. 62, 999-1002
(1958).
116. D. Cubicciotti and F. J. Keneshea, Jr., J. Phys. Chem. 63, 295 -
(1959).
-
U.7'.
.
F. J. Keneshea, Jr., and D. Cubicciotti, J. Phys. Chem. 63, lll2-15
(1959)
118. F. J. Keneshea, Jr., and D. Cubicciotti, J. F'hys. Chem. -
63, 1472-74
(1959).
119. -
M. A. Bredig, J. Phys. Chem. 63, 978 (1959).
L. E. Topol, S. W. Mayer, and L. D. Ransom, J. Phys. Chem. -
120.
(1960) . 64, 862

121. -,
J. D. Corbett, J. Phys. Chem. 62 l l 4 9 (1958).
122. -
J. D. Corbett, J. Am. Chem. SOC. 60, 475740 (1958).
123. L. E. Topol, S. J. Yosim, and R. A. Osteryoung, J. Phys. Chem. 65,
15ll (1961).
124. L. E. Topol and R. A. Osteryoung, "Electromotive Force, Polaro-
graphic, and Chronopotentiometric Studies i n Molten BismutbBismuth
Tribromide Solutions," J. Phys. Chem. ( t o be published).
c
57

125. C. R. Boston and G. P. Smith, J. Phys. Chem. -66


9
ll78-81 (1962).
-
126. A. H. W. Aten, Z. Physik. Chem. 66, 641 (1909).
-
127. A. H. W. Aten, Z. Physik. Chem. 73, 624-37 (1910).

. -
128. S. J. Yosim and E. B. Luchsinger, Ann. N.Y. Acad. Sei. 79, 1079-87
(1960)

i
c
59
I 3

,bi ORNL-3391
i UC-4 - Chemistry
1
i TID-4500 (20th ed., Rev.)

INTERNAL DISTRIBUTION

1. Biology Library 73. H. W. H 0 f f b . n


2-3. Central Research Library 74. C. A. Horton
4. Reactor Division Library 75. D. W. James
5-6. -
ORNL Y-12 Technical Library 76. G. H. Jenks
Document Reference Section 77. J. S. Johnson
7-41. Laboratory Records Department 78. S. S . K i r s l i s
42. Laboratory Records, ORNL R.C. 79. C. E. Larson
d
43. C. F. Baes, Jr. 80. A. P. Litman
44. C. J. Barton 81. C. H. Liu
45. P. B. Bien 82. H. F. McDuffie
46. R. E. Biggers 83. W. J. McDowell
47. F. F. Blankenship 84. A. S. Meyer
48. C. R. Boston 85. G. E. Moore
49-58. M. A. Bredig 86. R. L. Moore
59. J. H. Burns 87. J. E. Savolainen
60. W. L. Carter 88. C. S . Shoup, Jr.
61. G. H. Cartledge 89. M. J. Skinner
62. G. W. Clark 90. G. P. Smith
63. W. E. Clark 91. J. A. Swartout
64. E. L. Compere 92. E. H. Taylor
65. T. F. Connolly 93. R. E. Thama
66. W e H. Cook 94. W. C. Thurber
67. F. L. Culler 95. A. M. Weinberg
68. J. E. Eorgan 96. J. Pe YO-
69. H. A. Friedman 97. T. H. Davies (consultant)
70. R. A. Gilbert 98. D. F. Hornig (consultant)
71. G. Goldberg 99. I. P e r h a n (consultant)
72. Gerald Goldstein 100. J. D. Roberts (consultant)

EXTERNAT, DISTRIBUTION

101. Research and Development Division, AEC, OR0


102-685. Given d i s t r i b u t i o n as shown i n TID-4500 (20th ed., Rev.) under
Chemistry category (75 copies - CTS)

You might also like